"pilot projet admin burdens"
PILOT PROJECT ON ADMINISTRATIVE BURDENS Prepared by WiFo and CEPS for the European Commission, DG Enterprise under Contract Nr. B2/ENTR/05/091-FC FINAL REPORT TEAM: Michael Boeheim (Coordinator) Andrea Renda Hannes Leo Tom Weijnen Fabian Unterlass Paola Coletti Margit Schratzenstaller Vienna/Brussels, 14 December 2006 II TABLE OF CONTENTS 1 Introduction 1 1.1 Methodological note 3 1.2 Background Description of the prototype Standard Cost Model (SCM) 3 1.3 How to estimate the administrative burden using the SCM 7 1.4 Phases of a standard cost analysis according to the SCM Network 12 1.5 Countries that have adopted the SCM 14 1.6 The EU better regulation and simplification agenda 17 1.7 The proposed EU SCM 20 1.7.1 A comparison between the prototype SCM and the EU SCM 25 1.7.2 Pending issues after the pilot phase 27 2 Country case studies 29 2.1 The Netherlands 29 2.1.1 The Standard Cost Model 30 2.1.2 Baseline measurement 31 2.1.3 Methodology 32 2.1.4 Targets 44 2.1.5 Results 45 2.1.6 Most burdensome EU legislation 46 2.1.7 Conclusion 49 2.2 Denmark 51 2.2.1 Methodology 51 2.2.2 Content 61 2.2.3 Procedure 62 2.2.4 Targets 64 2.2.5 Results 66 2.2.6 Summary 74 2.3 The United Kingdom 77 2.3.1 Methodology 79 2.3.2 Procedural/organisational issues 86 2.3.3 Results 92 2.3.4 Most burdensome EU Regulations 93 2.3.5 Summary and comments 94 2.4 The Czech Republic 96 2.4.1 Methodology 96 2.4.2 Content 96 2.4.3 Procedure 97 2.4.4 Targets 101 2.4.5 Results 102 III 2.4.6 Most burdensome Directives 102 2.4.7 Summary – Lessons 103 3 Cross-country comparison and main findings 105 3.1 Cross-country comparison 105 3.2 Priority areas for simplification 113 3.3 Most burdensome EU directives 118 4 Suggestions for the Commission’s “Action Programme” 120 4.1 Methodological issues 122 4.1.1 Classification of origin 123 4.1.2 Database interoperability 128 4.1.3 Sampling, data collection, extrapolation 130 4.1.4 Level of detail 132 4.2 Organisational/Institutional issues 132 4.2.1 Scope and purpose 132 4.2.2 ‘Who should do what’ in the Action Programme 133 4.2.3 Institutional communication issues 135 4.2.4 Involving all EU institutions 136 4.3 Operational suggestions 137 4.3.1 Priority areas 137 4.3.2 Target-setting 139 4.3.3 Periodic reviews v. ‘live’ baseline and the role of Impact Assessment 141 4.4 Summary of main findings 142 IV LIST OF TABLES Table 1 – Examples of IOs .....................................................................................................9 Table 2 – Examples of administrative activities.................................................................10 Table 3 – The phases of a prototype standard cost analysis.............................................12 Table 4 – Overview of measurements for businesses per country...................................17 Table 5 – The phases in the Commission operational manual for ex ante assessment of administrative costs....................................................................22 Table 6 – Prototype SCM v. Commission operational manual.........................................24 Table 7 – EU standard reporting sheet...............................................................................25 Table 8 – Comparison between the Dutch SCM and the EU SCM................................... 26 Table 9 – Total Administrative Burden of the baseline measurement at 01-01- 2003, by ministry according to the ABC classification*) (€ million) ...............33 Table 10 - Example – selection of sample firms.................................................................39 Table 11 – Who does what in the 15 steps of the SCM, The Netherlands........................40 Table 12 – Realised net reduction and prognoses by Ministry (€ million)*) ...................46 Table 13 – Overview of results by department, domain and origin (€ million)*) ...........48 Table 14 – Who does what in the 15 steps of the SCM, Denmark ....................................64 Table 15 - Simplification potential in the three measured areas:......................................65 Table 16 – Central figures from the zero point and baseline measurements ..................66 Table 17 – Administrative burdens by areas and ABC-classification ..............................67 Table 18 – Top-5 most ‘EU-burdensome’ ministries/areas in Denmark .........................68 Table 19 – Results of the Danish measurement by ministry.............................................70 Table 20 – Most burdensome EU Directives in Denmark.................................................71 Table 21 - Multipliers for the entire economy and individual industries.......................73 Table 22 – number of regulations by department .............................................................82 Table 23 – UK reporting format on the origin of regulation (A and B)............................85 Table 24 – Who does what in the 15 steps of the SCM - UK.............................................87 Table 25 – Measurement methods ......................................................................................90 Table 33 – Who Does What in the 15 Steps of the SCM – Czech Republic .................... 100 Table 34 – Most burdensome EU directives in the Czech Republic............................... 103 Table 28 – Who does what – cross-country comparison................................................. 107 V Table 29 - Comparison of SCM variants .......................................................................... 109 Table 30 - Irritating areas for SMEs and their predominant origin according to IPAL................................................................................................................... 114 Table 31 – List of priority areas......................................................................................... 116 Table 32 – most burdensome EU directives, cross-country............................................ 119 Table 33 – Options available for the classification of origin ........................................... 127 Table 34 – Example of country distribution (CPB, 2006) ................................................ 131 Table 35 – Example: ‘who should do what’ in the Action Programme ........................ 134 Table 36 – Estimated total administrative burden for EU countries, % of GDP, 2003 .................................................................................................................... 140 LIST OF FIGURES Figure 1 – Administrative Burden versus Administrative Cost.........................................4 Figure 2 – Information obligation, data requirements and activities.................................5 Figure 3 – Decision tree for classification of origin .............................................................7 Figure 4 – Identifying the normally efficient business......................................................11 Figure 5 – Example of reporting sheet ...............................................................................14 Figure 6 – Decision tree for classifying data requirements - process...............................54 Figure 7 - Decision tree for classifying data requirements - contents..............................55 Figure 8 – The Danish Database model..............................................................................60 Figure 9 – Description of the Copenhagen Economics Trade Model (CETM) ................72 Figure 10 – Timetable of the upcoming Action Programme .......................................... 120 Figure 11 - Building the 'EU' Category............................................................................. 125 PILOT PROJECT ON ADMINISTRATIVE BURDENS EXECUTIVE SUMMARY Over the past few years, many European member states have launched policy initiatives aimed at measuring and reducing unnecessary administrative burdens caused by national and international legislation. The European Commission will be soon launching an Action Programme to measure and reduce administrative burdens generated by EU legislation in the EU25. A pivotal role in light of the upcoming Action Programme is played by the development of an EU common methodology to assess administrative burdens (the ‘EU SCM’), which should enable cross-country comparison and the identification of pieces of EU legislation which create unnecessary burdens on private and semi-private businesses, citizens, and public administrations. The reference model adopted by the Commission for the development of the EU common methodology is the Standard Cost Model (SCM) developed in the Netherlands. This model seeks to break down existing legislation into information obligations (IOs), which are further subdivided into data requirements (DRs) and administrative activities. Once the breakdown of administrative activities is achieved, each activity is given a time, cost and frequency variable to estimate the corresponding costs. A number of Member States adopted the Dutch model or developed their own national variant of the SCM. SCM adopters are however situated in different stages of applying the model. The Netherlands, Denmark, the United Kingdom and the Czech Republic already completed a baseline measurement, whereas others – including Austria, Belgium, France, Germany, Estonia, Hungary, Ireland, Italy, Norway, Poland and Sweden – are still carrying out the measurement or have just launched pilot projects on individual areas or pieces of legislation. The Netherlands as vanguard of the SCM prepared an ex-post inventory of total administrative burdens for the businesses in 2003. Under the lead of the Ministry of Finance (the Interministerial Project Unit for Administrative Burden (IPAL) is the central unit coordinating the implementation of the SCM, and is based in the Ministry of Finance) the measurement led to an estimated total for all administrative burdens of € 16.4 billion, as high as 3.6% of the Dutch GDP. Based on these results, the Dutch Cabinet outlined the target of reducing administrative burdens by 25% before the end of 2007. The Danish reduction target is similar to the Dutch one. Also in Denmark the reduction should be 25% of the already estimated administrative burdens. The difference is the timeframe for achieving the target, as the Danish government planned to achieve the target by 2010. The full scale measurement started in August 2004 and was completed in March 2006. The lead in the Danish measurement process was taken by the Division ii for Better Business Regulation in the Danish Commerce and Companies Agency (an agency under the Ministry of Economic and Business Affairs). All 15 ministries with business-relevant legislative and regulatory competences were involved in the measurement. The Danish approach also tried to investigate the trend of administrative burdens since November 2001 when the current Government was assigned. The total administrative burden was estimated as 2.1% of Danish GDP corresponding to 32.2 billion Danish Kroner. In the United Kingdom, the Better Regulation Executive (BRE) located within the Cabinet Office was established in May 2005 to coordinate, inter alia, the strategy for reducing administrative burdens. The Government had agreed to adopt the SCM two months before. The measurement started in September 2005 and the results are expected in 2006. Sixteen departments were involved in this measurement. In the meanwhile two separate exercises were carried out by the tax authority (HMRC) and the Financial Services Authority (FSA). Their results were published in March and April 2006. In the Czech Republic, the Better Regulation Unit, part of the Department of Regulatory Reform and Central State Administration Reform (located within the Office of the Government), was assigned to coordinate all activities targeting the reduction of administrative burden. An overall measurement was completed between March and December 2005. All business-related regulations in 13 ministries and 11 central administrative authorities were included. Of the other member states: § Sweden is already at the stage of carrying out the measurement process. Since July 2004 the Swedish Agency for Economic and Regional Growth (Nutek) is responsible for applying the SCM to measure administrative burden. First areas were measured in May 2004. In 2007 the overall measurement is expected to be completed. After the completion of each measurement the Swedish government sets a reduction target. § Estonia carried out two SCM pilot projects in 2005 to test the SCM. Up to now there is neither a reduction target set nor a SCM-related simplification project based on the pilot studies started. § Norway plans to measure administrative costs of businesses under the lead of the Ministry of Trade and Industry up to October 2007 to form the basis for considerable reduction targets by autumn 2009. § Italy started a pilot measurement whereas the SCM is being applied to 25 cases of permits and other administrative obligations for the exercise of business activities. Results are expected by September 2006. iii § After three pilot projects, Poland has just entered the initial phase of sector AB measurement, which will cover 150 information obligations identified in 20-25 legal acts.1 The results of the initial measurement will be presented at the end of 2006. § Ireland is interested in using the SCM and will therefore start pilot measurements in the coming months. § The Austrian government decided on 27 April 2006 to reduce the administrative burden for businesses by 25% until 2010. Therefore the SCM should be applied to estimate the administrative burden. Measurement results will be available in June 2007. § Germany has recently decided to carry out a full scale measurement exercise in 2007. § Finally, in France the last report of the Conseil d’État on the “Juridical security and complexity of the law” has identified the inflation of norms as a major cause of inefficiency and bad quality regulation. The Minister in charge of the reform of the State has launched a hundred modernisation audits, in order to alleviate the administrative burden for businesses and individuals. Moreover, at the end of 2005, some tests were carried out with external consultants on 110 authorisations concerning businesses, and over 30 concerning final users, so that a measure of administrative burden may be established. The tests shall be applied to all types of obligations, according to the SCM method.2 As a preliminary statement, it must be recalled that the scope and methodology of the SCM tools adopted by each of these member states differ noticeably. The SCM can be used both ex ante and ex post. When used ex ante, the SCM can be applied within regulatory impact assessment procedures, which helps preventing new pieces of legislation from creating unnecessary administrative burdens on businesses, citizens or administrations. The methodology applied in ex ante measurements is in some circumstances different from that used for ex post measurements. In addition, the integration of the SCM with RIA significantly depends on how well developed RIA is at national level. Countries that are measuring administrative burdens with SCM tools have targeted different areas, as shown in Table A below.3 The countries that have completed or plan to undertake a full baseline measurement are Austria, the Czech Republic, Denmark, Germany, the Netherlands, Norway and the UK. Other countries, such as Belgium and Poland, have planned to measure a significant portion of sectors. 1The project identified four burdensome areas: labour law, environment protection regulations, tourist services and public aid. 2 Conference on simplification, Paris, 9 June 2006. 3 Taken from SCM Annexes UK iv Table A – Overview of measurements for businesses per country Source: International SCM Network (2006) In addition, the European Commission has developed a proposed EU common methodology (“EU Net Cost Model”) since March 2005, has tested it in a pilot phase and published a revised version of the model in October 2005, now called ‘EU SCM’. An operational manual for applying the model was included in the Impact Assessment guidelines on 15 March 2006. The model has been applied in a number of Impact Assessments that have already been published or will be published by the end of this year. In this Report, we analyse the experience of the four member states that have fully implemented the SCM tool and completed a baseline measurement (The Netherlands, Denmark, the UK and the Czech Republic), with the aim of identifying the main features of these national models and comment on the compatibility between these models and the currently proposed EU methodology. The main finding of our analysis is that, although the models used in the four member states have a lot in common, many methodological and organisational differences exist. These differences, however, are not always relevant for the upcoming Action Programme: some methodological adjustments will in any case be needed, in order to allow the possibility of undertaking cross-country comparisons and benchmarking, as well as the effective communication between national and EU databases and their usefulness in highlighting potential areas for simplification. The most relevant differences that emerged from our case studies are the following: • Scope of measurement: some countries, such as the Czech Republic and Denmark, only calculate administrative costs faced by private enterprises, whereas the UK calculate costs for private and semi-private businesses, and the Netherlands also v include citizens. The EU model is the most ambitious, as it includes citizens and also public administrations. • Targets: most countries have selected a 25% ‘political’ target reduction of administrative costs. An exception is the Czech Republic, where the target selected is 20%; in addition, some countries – such as the Netherlands and, to a lesser extent, the UK – decided to set (at a later stage) specific targets for ministries/departments, whereas others – like Denmark and the Czech Republic – have not set different targets for specific ministries. • One-off costs: most countries – with the exception of the Czech Republic – do not include one-off costs in the baseline measurement. However, in the UK model one- off costs are included in the overhead, and the UK and Denmark include one-off costs in the ex-ante measurement carried out within RIAs. The proposed EU model specifies that such costs, when non-marginal, ‘may be taken into account’. This was also the approach chosen for the Dutch baseline measurement.4 In the Netherlands one-off costs are qualitatively described when ministries report the results of their measurement. • Obligations set by optional schemes: these information obligations are commonly referred to as ‘voluntary obligations’. In Denmark and the UK, these IOs are taken into account, if they are expected to arise from a quasi-regulation. In the UK, the baseline includes also codes of conduct and guidance documents with government backing, which cannot be considered as having binding force. • Distinction between ‘pure’ obligation and good practice written in the law: ‘Pure’ obligation refers to what one would stop doing if the legal obligation was removed.5 The EU model assesses all legal obligations even when the latter correspond to good practices. In other models, it is possible to assess only pure obligations. • Overhead: in the UK, a 30% rate was chosen, compared to the 25% rate used in the Netherlands and in Denmark. In the Czech Republic, a specific overhead percentage has been set by particular ministries, and no standard overhead percentage for all ministries and administrative authorities is foreseen. • Demarcation: most countries use a default 50:50 split whenever an IO/DR is found to fall under the competency of more than one department. In the UK, the need for demarcation was foreseen, but no 50:50 split rule was chosen. The most appropriate split is therefore set through negotiation between departments, and normally the department that has taken the regulatory initiative keeps ownership of the corresponding obligations. 4 See Manual, December 2003, p.15. 5This definition is comparable to the definition of "administrative burdens" (pure obligation) and its distinction from administrative costs (good practice) used in Figure 1 vi • Segmentation: in practice, segmentation criteria differed noticeably. In the UK, industry sector and type of activity proved more useful than firm size (except for the tax measurement undertaken separately). In the Netherlands, Denmark and the Czech Republic firm size was a guiding principle. • Accuracy and costs: the Danish and UK models seem likely to reach a higher level of accuracy in the estimates, also due to the intensive involvement of consultants, the more detailed breakdown of IOs into data requirements (Denmark) and the further breakdown of the classification of origin (Denmark and UK). The UK database seems the most suited to retrieve original pieces of EU legislation, although with some gaps. The greater accuracy, however, is reflected also in the cost of the measurement, especially as regards the UK. • Organisational patterns: the four national models exhibit some differences as regards the role played by ministries, central coordinating units and consultants. The Dutch and Czech models rely on ministries much more than the Danish and UK models. In the UK, ministries/departments are involved only in the initial step of the measurement, which is then carried out mostly by the consultant with the supervision of the BRE. In the Netherlands, IPAL is involved mostly in the review of activities performed by ministries and consultants, and in the final stages of the measurement. In the Netherlands, consultants are not involved in the final reporting and transfer to database, whereas in Denmark and the UK the setting up of a database was achieved with a strong involvement of consultants. The analysis of the four case studies allowed us to draw two main conclusions. On the one hand, the adoption of a common theoretical framework – provided by the SCM tool – certainly represents a significant advantage for the implementation of a multi- level measurement model coordinated by the European Commission. Had existing models obeyed to differing theoretical assumptions and methodologies, such an exercise would certainly have proven impossible from the outset. At the same time, it must also be observed that the differences in the four models analysed makes cross- country comparison a rather challenging exercise. In this Report we identify the most burdensome EU pieces of legislation according to the measurements carried out in the four surveyed member states, and develop a list of the ‘top-15’ (most burdensome) EU directives. In addition, we report the findings of the four central units in charge of coordinating the implementation of the SCM tool as regards possible priority areas for simplification. Finally, we offer a set of suggestions for the upcoming Commission Action Programme on administrative burdens, with specific focus on pending methodological and organisational issues. Table B below provides examples of burdensome pieces of EU legislation in each of the selected priority areas and illustrates the available information as regards the relative percentage of total administrative costs accounted for by each area in Category A and B in the four member states analysed. vii Table B – List of priority areas viii ix Table C lists the most burdensome EU directives, based on data from the central coordinating units in the four member states analysed.6 Table C – most burdensome EU directives, cross-country 6 Figures reported as ‘not available’ in Table C above do not necessarily correspond to pieces of legislation that were not measured by member states. Lack of availability of an exact figure for a given piece of legislation is often due to time constraints, or to the impossibility of tracing back the original piece of legislation in the national database. – x – After this Pilot Project, it will be possible to identify a list of low-hanging fruits based on informal proposals by the central units in charge of better regulation in the four member states analysed. These proposals are being prepared by the central coordinating units in the four member states analysed and will be filed as informal and unofficial contributions by these agencies, not to be interpreted as national or governmental views. This Final Report also provides suggestions for the organisational and methodological arrangements that should be adopted in order to effectively set up a pan-European measurement of administrative burdens. In this respect, the identified areas for optimisation of the current EU SCM are the following: 1. Database interoperability: to ensure the interoperability of national databases on administrative burden and access for the Commission. 2. Country distribution: identification of weighting systems for assessing EU-wide costs on the basis of a limited number of national data (e.g. country distribution). 3. Accuracy: identification of the average margin of error of administrative cost assessments. 4. Standard ratios: identification of standard ratios for overheads, training costs and learning curves and for costs corresponding to normal business operation, among other things. 5. Thresholds: identification of specific threshold(s) below which quantification is not necessary (minimum thresholds for the application of the model). 6. Extension to citizens: possible adjustments of the model when assessing administrative costs put on citizens. 7. Guidance on borderline cases: possible problems in distinguishing IOs from other regulatory costs and how to overcome them. 8. Standardisation of IOs and target groups: looking at possible shortcomings of the typologies of IOs and required actions used in the EU operational manual; examining the need for a typology of target groups. 9. Exchange of data: Organising optimal exchange of data between the Member States (including their regional authorities) and the Commission. 10. Target-setting: the issue of target-setting still needs to be addressed and agreed upon. Whether an overall target can be set at pan-European level for the reduction of administrative burdens imposed by legislation within a given timeframe is one of the issues that will have to be explored in the near future. This Report offers suggestions on these and other pending issues. In particular, we reach the following main conclusions. • Based on national experiences and the similarities between the EU and national models, launching a pan-European Action Programme is feasible, although challenging. • The issue of database interoperability and institutional communication between the EU and member states would be more easily solved if the Commission identifies all the IOs/DRs generated by the acquis, plus those that represent the ‘minimum efficient implementation’ of pieces of EU legislation. This activity – which leads to defining the – xi – ‘EU’ category (or category ’zero’) should be carried out by consultants in the first months of the Action Programme. Once the Commission has published a list of IOs/DRs, member states that already have a database should update it accordingly. Those that do not have a database yet should design it in order to allow for interoperability – i.e. by ensuring that all the IOs defined by the Commission are flagged and measured. • With this arrangement, member states will be called to validate Category ‘EU’ and then define their category B in a residual way, thus also identifying instances of gold-plating. • Member states are more suited to measure IOs/DRs provided by the Commission as ‘EU’. As most member states will have to carry out their own measurement anyway, such an arrangement increases economies of scope and avoids redundancies such as, eg, double interviews or duplication of focus groups, etc. Depending on the number of countries that participate to the measurement, the Commission will then have to use a country distribution to extrapolate the results to the EU level. Such exercise has already been undertaken in recent Commission RIAs that integrate the SCM tool. • If the Commission defines category ‘EU’, there would be no need to invest efforts and resources to build a ‘Table of Concordance’ between EU and national legislation during the preparatory phase of the measurement. At the end of the measurement, if member states have flagged the origin of – and the national references for – all IOs defined by the Commission as ‘EU’, the Table of Concordance would be automatically generated. • Assumptions made during the Action Programme and in extrapolating data to the EU level can then be standardised for use in ex ante impact assessment, though the SCM methodology for ex ante impact assessment (like occurs in the UK) will inevitably differ from that applied in the ex post measurement exercise. In ex ante measurement, as a matter of fact, one-off costs will always be taken into account, and ‘business as usual’ costs will never be considered. • As regards accuracy and the level of detail, we suggest – based on national experiences analysed – that the Commission identifies ‘IOs’ as the basic unit of its measurement, though subdividing IOs into DRs only in exceptional circumstance. Further subdivisions would add little accuracy, but a lot of complexity. Furthermore, the typical ‘Pareto distribution’ of administrative burdens suggests that in most cases, for IOs that are not particularly costly direct assessment of administrative cost is more efficient and effective than interviews or focus groups. • In operational terms, we suggest that the Commission awards priority to the 10 most burdensome areas in Categories A and B, after submitting them to consultation in the October 2006-February 2007 timeframe. These areas always represent at least 70% of burdens generated by EU legislation. The measurement of remaining areas and of burdens faced by citizens and public administrations should be undertaken at a later stage or, in any case, separately. • As the EU impact assessment system is becoming more established and pervasive in the Commission Strategic Planning and Programming Cycle, we suggest that the Commission opts for keeping a ‘live’ baseline. This requires some refinement in the – xii – current IA procedure: in particular, when undertaking ex ante IA, DGs, the Parliament and the Council should ensure that the IOs corresponding to new proposals are clearly identified, and that IOs that are repealed are also listed. • Targets can be set for priority areas. This does not preclude the potential for announcing an overall ‘political’ target, as was done in some of the analysed countries. After the measurement, targets can be adjusted for each DG. • When setting targets, it is essential that other EU institutions commit to cooperate with the Commission in achieving the targets set. Otherwise, simply setting targets for each DG would prove impossible. Such a coordination between the Commission, the Council and the Parliament should be sought through an Inter-Institutional Agreement. • Institutional communication between the Commission and member states can be achieved by replicating the fruitful collaboration with the International SCM Network that made this Final Report possible. As the participation of EU member states to the Network is increasing, this would automatically ensure that the Commission avails of a growing number of counterparts and data sources when undertaking the measurement. Moreover, the use of web-based tools such as SINAPSE would prove essential. PILOT PROJECT ON ADMINISTRATIVE BURDENS FINAL REPORT 1 INTRODUCTION The European Commission will be soon launching an Action Programme to measure and reduce administrative burdens generated by EU legislation in the EU25. A pivotal role in light of the upcoming programme is played by the development of an EU common methodology to assess administrative burdens, such as to enable cross- country comparison and the identification of pieces of EU legislation which create unnecessary burdens on private and semi-private businesses, citizens, and public administrations. The reference model adopted by the Commission for the development of the EU common methodology is the Standard Cost Model developed in the Netherlands and currently applied – although with some important methodological and procedural differences – in a growing number of member states with the aim of measuring and reducing administrative burdens. The reduction of administrative burdens is directly linked to the Lisbon strategy and the need to boost European competitiveness in the years to come, as confirmed by a number of recent studies. For example, in 2004 CPB, the Dutch Bureau for Economic Policy Analysis, estimated that the magnitude of administrative burdens reaches at least 3.6% (i.e. the Dutch percentage) of the EU GDP, i.e. 340 billion euros.7 Reducing such burden by 25% would imply, according to such estimates, an increase in GDP of 1.7%. In Denmark, similar studies show that a reduction in administrative burdens of 25% would result in a rise in Danish GDP of more than €1.5 billion (0.8%). In Denmark, the magnitude of administrative burdens has been estimated at 2.1% of GDP.8 Over the past few years, the European Commission has already devoted significant attention to the variants of the Standard Cost Model currently implemented in these member states. In early March 2005, the Commission presented a prototype model ("EU Net Administrative Cost Model") and announced its intention to put it to the test.9 On 22-23 March 2005, the European Council requested “the Commission and the Council to consider a common methodology for measuring administrative burdens with the aim of reaching an agreement by the end of 2005”, by taking “advantage of 7See P. Tang and G. Verweij, Reducing the administrative burden in the European Union, CPB Memorandum, 24 August 2004. See also below, at Section 4, for a more detailed discussion of these calculations. 8 Cfr. The Danish Reform Strategy. Contribution to EU’s Growth and Employment Strategy (Lisbon Strategy), October 2005. 9“Detailed outline of a possible EU Net Administrative Cost Model”, annexed to the Communication on Better Regulation for Growth and Employment in Europe , SEC(2005)175 of 16 March 2005. – 2 – the results of the Commission’s pilot projects”.10 The Commission tested the proposed methodology between April and September 2005. The pilot phase, on the one hand, confirmed the feasibility and added value of a common EU methodology, and on the other hand, suggested a number of amendments to the prototype. As a result and in line with the European Council's mandate, the Commission proposed in its Communication of 21 October 2005 the adoption of the “EU Net Cost model”, also called EU SCM.11 It listed a number of possible improvements to the EU SCM, while making clear that such optimisation was no precondition for its application. This was demonstrated by the decision to include the proposed methodology in the Commission’s impact assessment guidelines and evaluation guidelines (an operational manual for applying the model was included in the IA guidelines on 15 March 2006).12 In this Communication, the Commission also declared its longer-term intention to explore whether the proposed EU common methodology could be used to assess cumulative administrative burden at sectoral level. In this Communication, adopted on 21 October 2005, the Commission also declared its longer-term intention to explore whether the proposed EU common methodology could be used to assess cumulative administrative burdens at sectoral level. The main objectives of our analysis are: a) to identify policy areas that are most responsible for the creation of administrative burdens, based on results already obtained in the four Member States analysed; b) to explore how the SCM can best be applied for cross-country measurement; c) to suggest refinements in the EU SCM to allow its use for ex-post measurement. In this Interim Report, we focus on the variants of the Standard Cost Model adopted in the four member states (The Netherlands, Denmark, UK and the Czech Republic) that have completed the baseline measurement of administrative burdens to date. The results of this pilot project will be another step to further the Commission’s strategy for reducing administrative costs and setting common cost reduction targets. The pilot project’s results are also meant to lay the foundations of the Commission’s Action Programme for reducing administrative costs, to be published at the beginning of 2007. The Action Programme will launch a large measurement project across the European Union that will identify what legislation is most eligible for administrative costs reduction. This Final Report is structured as follows. In this section, we provide a background description of the SCM tool, including information on the countries that have adopted the SCM for the measurement of administrative costs at national level. We also provide 10 Point 24, Presidency conclusions 11 See Staff Working Document, Developing an EU common methodology for assessing administrative costs imposed by EU legislation - Report on the Pilot Phase (April– September 2005), SEC(2005) 1329, 21 October 2005 annexed to the Communication on a “EU common methodology for assessing administrative costs imposed by legislation”. 12These documents are available at: http://ec.europa.eu/governance/impact/docs_en.htm. The Commission pursued the optimisation of the methodology with the help of the Member States (Standard Cost Model Steering Group). In July 2006, it set up a virtual network of experts through SINAPSE, a web-platform for the collection of expertise. This will prepare the ground for the work of the High Level Group of national experts on better regulation, set up to advice the Commission on this issue and others. One aim of this is to agree on database interface and standard ratios for overheads. – 3 – information on the relevance of the Commission’s strategy for reducing administrative burdens within the more general Commission’s Better Regulation action plan. Section 2 contains a detailed description of national experiences in the four countries that have completed the baseline measurement, covering the methodological, procedural and organisational features of the models adopted, plus the results obtained in the baseline measurement. In section 3 we offer an assessment of the main similarities and differences in the models adopted by each of the surveyed member states and the proposed EU Net Cost Model. We also describe the findings of the pilot project as regards potential priority areas for simplification at EU level. This includes the identification of sectors most affected by administrative burdens and of the most burdensome EU directives and regulations. Section 4 provides some suggestions on methodological and organisational arrangements that will have to be adopted in light of the upcoming pan-European measurement. We also offer operational hints and suggestions on target-setting before and after the baseline measurement. 1.1 Methodological note This pilot study was carried out in the period July-October 2006 on the basis of materials and documents provided by the European Commission’s DG Enterprise and Secretariat General, as well as document and measurement results provided by officials of the surveyed member states that are in charge of implementing the SCM. In order to describe in detail the experience of the four member states, a questionnaire was sent to all four competent bodies and subsequent visits were organised to refine the main findings. Desk research was also undertaken to describe the main differences between the variants of the SCM adopted in each of the four member states, as well as to explore the main challenges for the full implementation of the EU SCM in the forthcoming Commission Action Programme on the reduction of administrative burdens. The authors are very grateful to officials in DG Enterprise and the Secretariat General of the European Commission, as well as to officials at the Dutch IPAL, the UK Better Regulation Executive, the Danish Commerce and Company Agency and the Czech Better Regulation Unit for providing valuable support and cooperation in the course of this study. 1.2 Background Description of the prototype Standard Cost Model (SCM) In this section we provide a description of the essential features of the Standard Cost Model (SCM), to be taken as reference for the comparative analysis that will be carried out in the next sections of this interim report. The SCM was developed by the Dutch Ministry of Finance to provide a simplified and consistent method to measure and consequently reduce the impact of legislative regulation on businesses – the so-called administrative burdens.13 The starting point is that regulation may impose costs – 13Countries exercising the SCM model are using slightly different definitions of businesses. In principle the term “businesses” means organisations in the private sector, whereas this definition could also include for example charities and the voluntary sector. – 4 – direct financial costs (taxes, fees, charges, penalties etc.) or costs of compliance with regulatory requirements – and accordingly hampers the competitiveness of businesses. To be able to contribute to competitiveness by reducing administrative burdens, the impact of legislation has to be measured. As a result, the SCM targets the quantitative dimension of administrative burdens. As shown in Figure 1 below, administrative burdens are considered as a part or subset of administrative costs. They are the part of administrative costs which is caused by regulatory requirements. Conversely, administrative costs normally also include activities which would be carried out also absent regulation. Figure 1 – Administrative Burden versus Administrative Cost Overall admnistrative costs Overall admnistrative costs Administrative costs Administrative costs Business Business from central from central administration costs administration costs government regulation government regulation Administrative activities that Administrative activities that Administrative activities that Administrative activities that businesses only conduct businesses only conduct businesses may continue if businesses may continue if because regulation requires because regulation requires the regulations were the regulations were it, i.e. administrative it, i.e. administrative removed removed burdens burdens Source: SCM Network, International Standard Cost Model Manual, p. 7 The main strength of the SCM is the high level of detail it can reach in measuring administrative costs through investigation of individual activities. The methodology followed by the SCM makes it possible to produce standardised figures for the resources used by businesses in order to comply with specific laws and executive orders. In practice, the SCM aims at identifying those textual parts of regulation that require businesses to make information available to public authorities or third parties. These textual parts are called ‘information obligations’ (IOs). It is possible – although often difficult – to subdivide these information obligations into smaller pieces called ‘data requirements’ (DRs). To fulfil the required information obligations – or rather, to produce the requested information – affected businesses normally have to carry out additional administrative activities. The costs of these additional activities may arise from internal consumption in form of use of employees’ time or on the other hand from external consumption of resources (e.g., fees for external experts, outsourcing costs, and cost of acquisitions). Therefore the administrative costs of a piece of legislation are defined as the costs of carrying out the various activities required by regulation. Figure 2 shows how the SCM splits the requirements of regulation into detailed activities, which can be measured or further estimated. – 5 – Figure 2 – Information obligation, data requirements and activities Regulation Internal costs Internal costs Information obligation 1 Information obligation 1 Data requirement 1 Data requirement 1 Activity 1 Activity 1 •Hourly rate •Hourly rate •Time •Time •Overhead •Overhead Activity 2 Activity 2 Data requirement 2 External costs External costs Data requirement 2 •Hourly rate •Hourly rate Information obligation 2 Information obligation 2 •Time •Time Activity n Activity n Acquisitions Acquisitions Data requirement n Data requirement n (Monetary value) (Monetary value) Information obligation n Information obligation n Source: SCM Network, International Standard Cost Model Manual, p. 9 This detailed breakdown of administrative costs enables the measurement of the anticipated administrative consequences of a draft piece of legislation before its implementation (ex-ante measurement), as well as of the factual administrative consequences for the businesses in respect of legislation already in force (ex-post measurement). Ex post measurements that are carried out on the overall administrative costs in a given policy area or on the entire corpus of legislation in force are defined as baseline measurements – i.e. a statement of the overall administrative costs that businesses must face in following a current set of regulations at a given point in time. Such measurements are also conducted in order to keep the baseline measurement updated over time with the consequences of new or amended regulations. The SCM also allows – in most cases – for differentiating the administrative burden according to sources of regulation. In the prototype SCM, the administrative burdens – more precisely, the DRs stemming from IOs contained in selected legislation – are segmented into burdens caused fully by international legislation; burdens caused by international legislation but the implementation of which is in the remit of national governments; burdens caused solely by national legislation; and burdens caused by regional or local legislation (for countries with a federal/regional structure). This methodology also makes it possible to further break down the first two categories by distinguishing between administrative burdens caused by the EU-legislation and those caused by other international legislation. However, the current implementation of the SCM in some EU member states, as will be explained in further detail in section – 6 – 2, currently does not allow for such an exercise.14 To the contrary, all surveyed member states distinguish between: § Category A data requirements: this category includes the data requirements (and their fulfilment) generated by legislation enacted at EU or international level. The related administrative burdens consequently fall outside the sphere of influence of national governments. § Category B data requirements: these include data requirements that arise from EU and international legislation and regulations, which allow for some degree of flexibility in the implementation at national level. As a result, although not originating from national legislation, there administrative burdens can be considered to fall at least partially under the sphere of influence of national governments. § Category C data requirements: this category includes data requirements generated by national legislation and regulations. The extent (and reduction) of the related administrative burdens can therefore be considered to fall fully under the sphere of influence of national governments. These categories do not include administrative burdens caused by legislation or regulations enacted at sub-levels of government (e.g. regional legislation). This is simply due to the fact that countries that have fully implemented the SCM to date do not exhibit a multi-level constitutional structure. However, other countries – e.g. Italy – are introducing an additional category (termed ‘Category D’) to account for regional legislation in implementing their variant of the SCM tool. According to the current practice in most countries that have adopted the SCM, the classification of regulatory origin is achieved through seven steps of analysis, that are reported below, in Figure 3. 14 Notable exceptions are Denmark and the UK. – 7 – Figure 3 – Decision tree for classification of origin Yes 1. Is the data requirement solely a result of national policy? 1. Is the data requirement solely a result of national policy? Category C Category C NO 2 Does the data requirement imposed in a European or international context specifically 2 Does the data requirement imposed in a European or international context specifically Yes state the way in which this obligation must be met? state the way in which this obligation must be met? Category A Category A NO 3. Does the data requirement imposed in a European or international context contain 3. Does the data requirement imposed in a European or international context contain national information obligations (that are not internationally obligatory)? Such as the national information obligations (that are not internationally obligatory)? Such as the imposition of extra requests for information. imposition of extra requests for information. Either // or Either or 4. Is the data requirement imposed in a European or international context expanded by the 4. Is the data requirement imposed in a European or international context expanded by the national government? Such as by increasing the frequency of information obligations or national government? Such as by increasing the frequency of information obligations or imposing the information obligation on additional target groups. imposing the information obligation on additional target groups. 5a. Classify the ‘national’ section of the data requirement in 5a. Classify the ‘national’ section of the data requirement in Yes Category C Category C Category C Category C Yes Category A Category A 5b. Are the means of implementation also stipulated in the 5b. Are the means of implementation also stipulated in the Yes ‘international’ section of the requirement ‘international’ section of the requirement NO NO Category B Category B 6. The data requirement arises from European or international directives but their 6. The data requirement arises from European or international directives but their implementation is not stipulated internationally. implementation is not stipulated internationally. Yes Either // or Either or Category B Category B 7. The data requirement is imposed by the national government for 7. The data requirement is imposed by the national government for implementation/enforcement of European or international legislation/regulation implementation/enforcement of European or international legislation/regulation Source: SCM Network, International Standard Cost Model Manual, p. 27. 1.3 How to estimate the administrative burden using the SCM As already mentioned, the basic idea in the SCM tool to estimate the administrative burden of regulation is to split observed pieces of legislation into information obligations, and further break down information obligations in data requirements. Each data requirement is then expressed in terms of administrative activities: the cost of each administrative activity is then estimated with the following basic formula: Cost per administrative activity = Price x Time x Quantity (population x frequency) Whereas: § the price is a tariff for additional activities, divided in internal – e.g. hourly wage as a measure of the unit costs of additional activities demanded by information obligations – and external prices – e.g. unit costs (hourly rates) of outsourcing the activity; whereas only activities demanded by the information obligation are relevant. – 8 – § Time means the number of time units (e.g. hours) needed to perform the required activity; § Quantity represents how often the activity has to be carried out per year (frequency) by the effected number of businesses (population). Additional costs (e.g. necessary acquisitions) also have to be considered as elements of cost relevant to the administrative activity at hand. Following this formula for each activity of an information obligation, the administrative costs generated by a piece of legislation can be estimated by summing up the costs of all activities and all information obligations caused by that legislation. Thus, the relevant structure of splitting-up regulation can be illustrated as in Figure 2 above. The abovementioned measurement exercise seems, at least conceptually, simple. In reality, problems may arise, for example, because of the need to identify relevant cost parameters and gather information on prices, time and quantities.15 As shown by vanguards of the SCM such as the Netherlands and Denmark, the best way to collect data seems to be asking the businesses themselves what they are doing or have to do to fulfil the required information obligations. Consulting a sample of affected businesses and also business experts can deliver the required data to estimate the costs for sectors and also for regulation areas. Such consultation can take different forms, from face-to- face or telephone interviews to the organisation of workshops or business panels. However, sampling patterns and choices can affect the reliability of the final estimates: for such reason, other tools can be used for data collection, ranging from the consultation of experts to ad hoc consultancy studies. The SCM provides for tools that are aimed at reducing the uncertainty related to the model implementation, such as standardised lists of IOs and administrative activities, and criteria for selecting the so- called “normally efficient firm”, which is taken as reference for the calculation of price, time and quantity in the formula above. As regards standardised lists, the Manual published by the International SCM Network to reduce administrative burdens offers a non exhaustive list of IOs and administrative activities. Table 1 below shows the list of IOs adopted as reference by the International SCM Network. 15Frequency is normally easier to determine, as mandatory fulfilment of some IOs is usually enshrined in the wording of the legislation. – 9 – Table 1 – Examples of IOs • Returns and reports: This relates to returning and reporting information, e.g. tax deducted from income at source. • Applications for permission for or exemption from…: This relates to all types of application for permission for or exemption from various activities, e.g. application for a licence to sell spirits. • Applications for authorisation: This relates to applications for authorisation to carry out certain activities, e.g. authorisation as a sewer contractor. • Notification of activities: This relates to businesses having to notify the authorities of specific activities, e.g. notification of the transportation of dangerous cargo. • Entry in a register: This relates to businesses having to be entered in a register or on a list, e.g. entry in the business register. • Carrying out inspections of…: This relates to the business itself carrying out inspections of machinery and equipment that can represent a risk to health or the environment, or monitoring the conditions for employees. Inspections are normally carried out by certified organisations, e.g. drawing up a workplace assessment. • Applications for subsidies or grants for…: This means the business applying for a subsidy or the like, e.g. a subsidy for job training. • Keeping commercial emergency plans and programmes updated, etc…: This relates to the business keeping those documents required by the authorities up to date. It would include manuals and emergency plans, for example. • Cooperating with audits/inspections of…: This relates to informing and assisting inspectors who carry out inspections of and auditing work for a business, or who visit a business in connection with enforcement of a regulation. • Statutory labelling for the sake of third parties: This means, among other things, labelling products or installations with consumer information, e.g. energy labelling of domestic appliances. • Providing statutory information for third parties: This relates to providing third parties with information (as distinct from labelling), e.g. a financial prospectus to accompany investment products. • Framing complaints and appeals: This relates to submitting complaints about and (possibly later) appealing against a decision made by the authorities. This information obligation should only be analysed if it is characteristic of a normally efficient business to complain in the area in question. Source: SCM Network, International Standard Cost Model Manual, p. 24. In addition, Table 2 contains a reference list of administrative activities that was developed by the International SCM Network. – 10 – Table 2 – Examples of administrative activities 1. Familiarisation with the information obligation. The resource consumption of businesses in connection with familiarising themselves with the rules for a given information obligation. 2. Information retrieval. Retrieving the relevant figures and information needed to comply with a given information obligation. 3. Assessment. Assessing which figures and information are necessary for the public authorities to accept the report. 4. Calculation. Performing the relevant calculations needed for the public authorities to accept the report. 5. Presentation of figures. Presenting the calculated figures in tables or the like. 6. Checking. Checking the calculated figures, e.g. by reconciliation with other data. 7. Correction. If the business’s own checks reveal errors in the calculations, corrections are made afterwards. 8. Description. Preparation of description, e.g. the directors’ report in the Danish Financial Statements Act. 9. Settlement/payment. Payment of tax, charges or the like. 10. Internal meetings. Meeting held internally between the various personnel groups involved in complying with the information obligation. 11. External meetings. Meetings held in cases where compliance with the information obligation requires meetings with an auditor, lawyer or the like. 12. Inspection by public authorities. Businesses must assist external inspectors when they carry out their inspection at the business. 13. Correction result from inspection by public authorities. If the external inspection identifies faults/defects, corrections are made afterwards. 14. Training, updating on statutory requirements. Relevant employees must be kept up to date with rules that change frequently (at least once a year). 15. Copying, distribution, filing, etc. In some cases the report is copied, distributed and/or filed in order to comply with the information obligation. It may also be necessary to store the information obligation with a view to subsequent production in connection with an inspection. 16. Reporting/submitting information. In cases where compliance with an information obligation requires the submission of information on the business, the information must be sent to the relevant authority. Source: SCM Network, International Standard Cost Model Manual, pp. 25-26. Finally, as regards the identification of the “normally efficient business”, the SCM tool is not conceived to account for burdens generated by the inefficiency of the businesses, but is meant to capture the impact that legislation in force might exert on businesses – 11 – that carry out their tasks in a reasonably efficient way. For such reason, once the data have been collected, e.g. through interviews or focus groups, the SCM prescribes that the consultant or official in charge of implementing the SCM proceeds to assess how long it would take a normally efficient business to carry out the various administrative activities required to comply with a data requirement for a given IO. On this basis, the internal/external/acquisition costs faced by the businesses to comply with the IO are calculated. Figure 4 below illustrates some examples of how collected data can be standardised in order to produce a reliable estimate for implementation within the SCM tool. Figure 4 – Identifying the normally efficient business Administrative activity A: Administrative activity B: Company 1 10 min. Company 1 10 min. Company 2 10 min. Company 2 20 min. 10 min. 15 min. Company 3 10 min. Company 3 10 min. Company 4 10 min. Company 4 20 min. Company 5 30 min. Company 5 15 min. Administrative activity C: Administrative activity D: Company 1 10 min. Company 1 10 min. Company 2 20 min. Company 2 20 min. More 20 min. Company 3 50 min. interviews Company 3 25 min. Company 4 2 min. Expert 1 20 min. Company 5 5 min. Expert 2 15 min. Source: SCM Network, International Standard Cost Model Manual, pp. 55 As explained by the International SCM Network, as regards activity A in Figure 4, business 5 is not counted as a normally efficient business, as it is too different from the others in the sample. In the case of activity B, the final figure is selected by the consultant or competent official, whereas for activity C, more interviews are needed in light of the significant variance of the results. In the case of activity D an expert assessment is made of what the level is for a normally efficient business.16 The reliability of such estimates – and the variance of the data collected – normally depend on the accuracy of the segmentation carried out prior to data collection. The more accurate the segmentation, the easier should be to reach a reasonable assessment of the time needed for a normally efficient firm to perform selected tasks. 16 See International SCM Manual, at 55. – 12 – This also means that one of the most crucial features of the SCM tool is the process that must be followed during the baseline measurement. The next section summarises the main phases of this process. 1.4 Phases of a standard cost analysis according to the SCM Network The SCM Network International Manual provides a description of the phases in which a standard cost analysis can or should be broken down. These are presented below, in Table 3. Table 3 – The phases of a prototype standard cost analysis Phase 0. Start-up • The business-related regulation to be included in the analysis is identified before the preparatory analysis is started. In the case of large analyses, especially baseline measurements, but also certain ex-ante analyses and updates, start-up meetings attended by the responsible ministry, the Commerce and Companies Agency, the consultants and any other relevant parties are held. Phase 1. Preparatory analysis Step 1 The text of the regulation is broken down into a number of information obligations. The information obligations are then broken down into data requirements and relevant administrative activities are identified. The responsible ministries then start to categorise the regulation into national and international elements. Step 2 Related regulation is identified and demarcation carried out. Step 3 The information obligations are classified according to type. Step 4 The relevant business segments are identified. Step 5 Identification of population, rate and frequency. Step 6 Clarification of which information obligations are to be assessed by means of business interviews and which will have time consumption estimated later in phase 3 (step 14) by means of expert assessment. Step 7 The relevant cost parameters are identified. Step 8 An interview guide is prepared for use in interviews. Step 9 An expert review of steps 1-8 is carried out. Phase 2. Analysis of time consumption and costs in businesses Step 10 Typical businesses from the legislation’s target group are selected for interview. Step 11 The businesses are interviewed. Step 12 Standardisation of time and resource consumption per activity per business segment. – 13 – Step 13 An expert review of steps 10 and 12 is carried out. Phase 3. Calculation and presentation Step 14 The validated data are extrapolated to a national level. Step 15 The results are presented in a report and the data set is entered in the overall database. Source: SCM Network, International Standard Cost Model Manual, pp. 20-21. As shown in Table 3, the analysis follows four main phases and requires the involvement of ministries/departments, consultants, stakeholders and (where available) a central coordinating unit. In phase 0 (“Start-up”), initial meetings of departments, consultants and other key stakeholders have to take place to clarify the areas to be investigated and carry out preliminary activities, such as preparing a data set of business-related regulation to be included in the analysis. In phase 1 (“Preparatory analysis”) decisions are made how to carry out the analysis. Therefore questions concerning the data requirement have to be clarified. One of the most important questions here is which business-related information obligations are chosen. After identifying the fields of interest (information obligations) the administrative activities done by the businesses to fulfil these obligations have to be specified. The next step is to identify the relevant background variables (e.g. the cost parameters) and finally the business segments affected by the regulation. An important aspect not only at this phase is to involve relevant experts in evaluating and validating the classifications and specifications done in the preparatory work of “Phase 1”. In phase 2 (“Analysis of time consumption and costs in businesses”) empirical data is collected from interviews with a sample of typical businesses belonging to the population affected by a given piece of legislation. Therefore a decision has to be made which businesses are chosen to be asked to cover all relevant segments in the population. The chosen businesses should also adequately represent the business structure in the sector of the national economy affected by the legislation at hand. As already recalled, there are different ways to conduct the interviews, including face-to- face, telephone or focus group interviews. After collecting the empirical data, phase 3 (“Calculation and presentation”) completes the process. At this stage, the collected data are scaled up to the national level for each segment and also for each administrative activity. Presentation of the measurement results implies that data reported show the most burdensome regulations and identify areas for reform. Further calculations demand the transfer of the data to a database. The reporting sheet is normally standardised: an example of reporting sheet is provided at Figure 5. – 14 – Figure 5 – Example of reporting sheet Source: SCM Network, International Standard Cost Model Manual, p. 59.. The SCM, as already recalled, is not only a tool for the measurement of administrative costs imposed by legislation, but also for the reduction of such costs. As a result, following the four main phases of the analysis, priority areas for reform are identified and targets are often assigned to competent authorities (e.g. ministries) to realise the potential for administrative simplification. 1.5 Countries that have adopted the SCM Since its development in the Netherlands, the SCM has been adopted by many other EU member states. These countries are situated in different stages of applying the model. The Netherlands, Denmark, the United Kingdom and the Czech Republic already completed a baseline measurement, whereas others – including Austria, Belgium, France, Germany, Estonia, Finland, Hungary, Ireland, Italy, Latvia, Norway, Poland and Sweden – are still carrying out the measurement or have just launched pilot projects on individual areas or pieces of legislation. The Netherlands as vanguard of the SCM prepared an ex-post inventory of total administrative burdens for the businesses in 2003. Under the lead of the Ministry of Finance (the Interministerial Project Unit for Administrative Burden (IPAL) is the central coordinating unit coordinating the implementation of the SCM, and is based in the Ministry of Finance) the measurement led to an estimated total for all administrative burdens of € 16.4 billion, as high as 3.6% of the Dutch GDP. Based on these results, the Dutch Cabinet outlined the target of reducing administrative burdens by 25% before the end of 2007. The Danish reduction target is similar to the Dutch one. Also in Denmark they reduction should be 25% of the already estimated administrative burdens. The – 15 – difference is the timeframe for achieving the target, as the Danish government planned to achieve the target by 2010. The full scale measurement started in August 2004 and was completed in March 2006. The lead in the Danish measurement process was taken by the Division for Better Business Regulation in the Danish Commerce and Companies Agency (an agency under the Ministry of Economic and Business Affairs). All 15 ministries with business-relevant legislative and regulatory competences were involved in the measurement. The Danish approach also showed the development of the level of administrative burdens since November 2001 when the current Government was assigned. The total administrative burden was estimated as 2.1% of Danish GDP corresponding to 32.2 billion Danish Kroner. In the United Kingdom, the Better Regulation Executive (BRE) located within the Cabinet Office was established in May 2005 to coordinate, inter alia, the strategy for reducing administrative burdens. The Government had agreed to adopt the SCM two months before. The measurement started in September 2005 and the results will be made publicly available in November 2006. Sixteen departments were involved in this measurement. In the meanwhile a separate exercise was performed by the tax authority (HMRC) and the Financial Services Authority (FSA) of the UK. Their results were already published in March and April 2006. In the Czech Republic, the Better Regulation Unit – part of the Department of Regulatory Reform and Central State Administration Reform (located within the Office of the Czech Republic Government) – was assigned to coordinate all activities targeting the reduction of administrative burden. An overall measurement was carried out between March and December 2005. All business related regulations in 13 ministries, 11 central administrative authorities were included. Examples from other member states include the following: § Sweden is already at the stage of carrying out the measurement process. Since July 2004 the Swedish Agency for Economic and Regional Growth (Nutek) is responsible for applying the SCM to measure administrative burden. First areas were measured in May 2004. In 2007 the overall measurement is expected to be finished. After the completion of each measurement the Swedish government sets a reduction target. § Estonia carried out two SCM pilot projects in 2005 to test the SCM. Up to now there is neither a reduction target set nor a SCM-related simplification project based on the pilot studies started. § Norway plans to measure administrative costs of businesses under the lead of the Ministry of Trade and Industry up to October 2007 to form the basis for considerable reduction targets by autumn 2009. § Italy started a pilot measurement whereas the SCM is being applied to 25 cases of permits and other administrative obligations for the exercise of business activities. Results are expected to be available by September 2006. – 16 – § After three pilot projects, Poland has just entered the initial phase of sector AB measurement, which will cover 150 information obligations identified in 20-25 legal acts.17 The results of the initial measurement will be presented at the end of 2006. § Ireland is interested in using the SCM and will therefore start pilot measurements in the coming months. § The Austrian government decided on 27 April 2006 to reduce the administrative burden for businesses by 25% until 2010. Therefore the SCM should be applied to estimate the administrative burden. Measurement results will be available in June 2007. § Germany has recently decided to carry out a full scale measurement exercise in 2007. § Finally, in France the last report of the Conseil d’État on the “Juridical security and complexity of the law” has identified the inflation of norms as a major cause of inefficiency and bad quality regulation. The Minister in charge of the reform of the State has launched a hundred of modernisation audits, in order to alleviate the administrative burden for businesses and individuals. Moreover, at the end of 2005, some tests were carried out with exterior consultants on 110 authorisations concerning businesses, and over 30 concerning the users, so that a measure of administrative burden may be established. The tests shall be applied to all types of obligations, according to the SCM method18. As a preliminary statement, it must be recalled that the scope and methodology of the SCM tools adopted by each of these member states differ noticeably. The SCM can be used both ex ante and ex post. When used ex ante, the SCM can be applied within regulatory impact assessment procedures, which helps preventing new pieces of legislation from creating unnecessary administrative burdens on businesses, citizens or administrations. The methodology applied in ex ante measurements is in some circumstances different from that used for ex post measurements. In addition, the integration of the SCM with RIA significantly depends on how well developed RIA is at national level. Moreover, countries that are measuring administrative burdens with SCM tools have targeted different areas, as shown in Table 4 below.19 17The project identified four burdensome areas: labour law, environment protection regulations, tourist services and public aid. 18 Conference on simplification, Paris, 9 June 2006. 19 Taken from SCM Annexes UK – 17 – Table 4 – Overview of measurements for businesses per country 1.6 The EU better regulation and simplification agenda The Better Regulation agenda of the Commission is deeply rooted in EU overarching legal principles, most notably in the application of the Treaty-based proportionality principle to be found in the EC Treaty, provides that the Commission should “take duly into account the need for any burden, whether financial or administrative, falling upon the Community, national governments, local authorities, economic operators and citizens to be minimised and proportionate to the objective to be achieved”.20 The current EU Better Regulation and simplification strategy can be traced back to the mandate issued by the European Council at the Lisbon meeting in 2000, and later confirmed at the Stockholm, Laeken and Barcelona summits. Under such mandate, the European Commission launched an extensive consultation and published an initial interim report (for the Stockholm Council), a White Paper on European Governance and a political communication (for the Laeken European Council), which ultimately led to the publication of the European Commission’s Action Plan on “Simplifying and improving the regulatory environment”, published on June 5, 2002.21 In that Action Plan, the Commission followed most of the recommendations issued by the high-level consultative group (the so-called “Mandelkern Group”) set up by the Ministers responsible for the Civil Service in November 2000. The Action Plan included actions falling within the Commission’s remit, actions proposed to the European Parliament and Council, actions designed to ensure the effective transposition and application of legislation by the Member States, and initiatives to promote a common legislative culture within the EU. 20 Protocol 30, on the application of the principles of subsidiarity and proportionality, Art. 5. 21See European Commission, Action Plan on Simplifying and improving the regulatory environment (COM(2002) 278), 2002. – 18 – Alongside the 2002 Action Plan, the Commission issued two communications, one on minimum standards for consultation and another which launched the new EU Integrated Impact Assessment (IIA) model.22 The objectives of the latter were to “improve the quality of Commission proposals, to ensure an analysis of the economic, environmental and social impacts of a proposal and to improve and simplify the regulatory environment”. Reference to the assessment of administrative burdens resulting from proposed new legislation was already included in the Technical Annex to the 2002 Impact Assessment Guidelines. However, no specific strategy on the measurement and reduction of administrative burdens caused by EU legislation was initially foreseen in the 2002 Action Plan, neither as a stand-alone tool, nor within the impact assessment model. In subsequent years, the Commission showed increased attention for the problem of administrative burdens both in terms of ex post screening of existing legislation23, and launching a pilot project on indicators of regulatory quality (IRQ) and a project for the ex post evaluation of the costs EU legislation and their burden on businesses, which led to the publication of a final report in May 2005.24 The Commission has developed a set of structural indicators in the framework of the Lisbon strategy. There is, however, no specific indicator on administrative obligations imposed by legislation. In May 2003, the Commission committed itself to “develop, in close cooperation with Member States, appropriate indicators to measure progress towards a higher-quality regulatory framework and lower administrative burdens, starting with the Internal Market”.25 The 2003 Inter-institutional Agreement on Better Law-making, the joint Letter of the Irish and three incoming presidents of ECOFIN filed on January 2004, the “Doorn Motion” within the Legal Affairs Committee of the Parliament and the Joint statement issued on December 2004 by the Irish, Dutch, Luxembourg, UK, Austrian and Finnish presidencies on “Advancing Regulatory Reform” were all evidence of a growing interest for linking impact assessment to competitiveness at EU level. In particular, in the response of President Prodi to the 2004 Joint Letter by the Irish and three incoming presidencies of ECOFIN, the Commission agreed to examine whether “assessment relating to the administrative burden for companies needs to be further improved”, acknowledging the relevance of this issue. The next day, the Council (ECOFIN) called for the development of a method for measuring the administrative burden on business, and the invitation was later confirmed in both the Spring European Council 2004 and the European Council on 4-5 November 2004, which supported the idea of developing of a common methodology for measuring administrative burdens, and welcomed the Commission’s intention to present a 22 See Renda, Impact Assessment in the EU. The State of the Art and the Art of the State, CEPS, 2006. 23 Framework action “Updating and simplifying the Community Acquis” (COM(2003)71, February 2003) 24 See Rambøll, Ex-Post Evaluation of EC Legislation and its Burden on Businesses, Study for European Commission DG Enterprise, May 2005; the final results of the IRQ project are available at http://www.bradford.ac.uk/irq/ (visited on 9 September 2006). 25 Communication from the Commission, Internal Market Strategy – Priorities 2003-2006, COM(2003) 238 final, 7 May 2003. Section B.8. Simplifying the regulatory environment, (b) Actions, No 4. – 19 – communication on such issue, and to refine the methodology with the cooperation of Member States through pilot projects before implementing the common methodology in its revised guidelines on impact assessment. On March 16, 2005 the Communication on Better Regulation for Growth and Employment in the European Union included, as a companion Staff Working Paper, a detailed outline of a possible “EU Net Administrative Cost Model” built with reference to the Dutch SCM. In the Staff Working Paper, administrative costs were defined as the “costs arising from reporting and information obligations laid down by legislation”, and considered as “one among several types of regulatory costs faced by businesses, public authorities and citizens”.26 In particular, in the Staff Working Paper the Commission assessed the merit of developing a common approach to the measurement and reduction of administrative burdens. The main expected benefits of such common approach were the following: § clarity about possible differences in procedures adopted at EU and member state level; § easier cross-country or cross-policy area comparison, benchmarking and the development of best practices; § economies of scale in terms of data collection and validation. However, the Commission clarified at the outset that it was not planning to apply the model within a wide measurement exercise covering entire sectors or even society as a whole: “preliminary analysis suggests that it would be difficult and not cost efficient to use an approach such as the ‘EU Net Administrative Cost Model’ … for estimating total administrative costs imposed on society or on selected sectors. The difficulty would come in particular from the wide variety of regulatory cultures in the Union, decreasing reliability of sectoral aggregates of micro-assessments, and the fact that applying the Dutch SCM at EU level would cost an estimated EUR 100 million”27 To the contrary, the Commission initially conceived its EU Net Administrative Cost Model as a tool to be applied to individual pieces of new (ex ante measurement) or existing (ex post measurement) legislation. In this respect, the EU Net Administrative Cost Model was meant to be integrated in the EU integrated impact assessment model, to enable the ex ante measurement and reduction of administrative burdens caused by proposed legislation; and at the same time, as a tool to measure the potential for cutting red tape in the ex post measurement of selected pieces of EU legislation. 26European Commission, Staff Working Paper, Annex to the 2005 Communication on Better Regulation for Growth and Jobs in the European Union, Minimizing Administrative Costs Imposed by Legislation, Detailed Outline of a Possible EU Net Administrative Cost Model, SEC(2005)175, 16 March 2005. 27 Id. (emphasis in original). – 20 – The outline of an EU Net Administrative Cost Model was then amended and refined through a pilot phase carried out from April to September 2005. The pilot phase was aimed at testing ways of assessing administrative costs imposed by EU legislation, and led the Commission to present a revised methodology, now included as Annex 10 to the Commission’s Impact Assessment Guidelines.28 The revised methodology, also called EU SCM, is described in detail in the next section. 1.7 The proposed EU SCM The European Commission’s proposed “common methodology for assessing administrative costs imposed by legislation” exhibits significant similarities but also important differences with respect to the Standard Cost Model described above, at Section 1.2. As far as similarities are concerned, the EU SCM confirmed the merit of adopting a “net cost” approach, whereas net costs are the costs introduced by legislation minus the costs eliminated by legislation at EU and/or national level. As confirmed by the Commission, this approach exhibits several theoretical and practical advantages over alternative approaches, in that it is consistent with the Commission’s impact assessment guidelines as well as with national RIA manuals, but it is also consistent with the approach adopted by member states that have completed the baseline measurement using the SCM tool. In addition, a net cost approach is especially useful in the ex ante measurement of administrative burdens, since it allows for a more precise assessment of the impact that the proposed piece of legislation will exert on administrative burdens. Furthermore, reported advantages of a net cost approach include the suitability for forthcoming sectoral estimates and assessments of cumulative burdens, and cost savings as regards database updating – a database built on net costs is constantly and automatically updated. Secondly, contrary to what the Commission had proposed in earlier documents, the EU SCM was found to be applicable to the assessment of cumulative burdens, and thus also for ambitious ex post measurement exercises encompassing whole sectors or – less likely – the entire corpus of legislation. In this respect, too, the new methodology can be said to be more comparable with those in use in countries that have adopted and implemented the SCM tool to date. In addition, in the pilot phase launched in April 2005 – which launched 3 pilot projects for ex post evaluation and 5 projects for ex ante evaluation29 – the Commission managed to use common definitions and reporting formats, thus achieving an important step forward in the direction of a common methodology to be applied by all DGs. The ex ante projects carried out, most of which have been completed in 2006, also helped the Commission in reaching wide consensus on the desirability of integrating the common methodology within the Commission’s 28 See Impact Assessment Guidelines, SEC(2005)791, 15 June 2005. 29 See Annex 2 to the Commission’s Staff Working Document, Annex t the Communication on a common EU methodology for assessing administrative costs imposed by legislation, Outline of the proposed EU methodology and Report on the Pilot Phase (April-September 2005), SEC(2005)1329, 21 October 2005. – 21 – impact assessment model. This, in turn, also means that the same methodology will be applied both for ex ante and ex post assessments, another feature which appears in line with the SCM tool. Thirdly, most of the most important methodological features proposed by the Commission exhibit strong similarities with the prototype SCM illustrated in section 1.2. above. For example, the Commission proposes to use the same core equation, the same relevant cost parameters (e.g. internal/external tariffs, cost of equipment and supplies) and the same formulas used to express the frequency of administrative activities, and the same approach to the assessment of the performance of a “normally efficient entity”.30 However, the EU SCM shares a number of peculiarities with national SCM variants. But also some important differences. These can be summarised as follows: § The EU common methodology appears ambitious as regards the type of administrative burdens that should be assessed, since it postulates that the model should account for burdens imposed by legislation on enterprises, the voluntary sector, public authorities and also citizens. As will be explained below, the EU common methodology is similar to the current version of the Dutch SCM in this respect, but appears broader in scope than other models, such as the Danish and UK models. § The EU common methodology includes not only information obligations towards public administrations (e.g. accounting requirements), but also information obligations towards private parties, such as consumers (e.g. labelling). In this respect, the EU SCM is similar to the Dutch and Danish model. § In the EU SCM, the burdens to be assessed include (non marginal) one-off costs, which is left as an option in the prototype SCM. § The EU SCM, like the International SCM Manual, differentiates between administrative costs and administrative burdens, but recommends assessing administrative burdens and leaving the issue of administrative costs to qualitative analysis. § The proposed methodology distinguishes clearly between ‘voluntary’ and ‘compulsory’ obligations, and excludes the former from the burdens to be assessed. § In seeking compliance with the principle of proportionate analysis, the EU methodology awards particular importance to the definition of the most suitable thresholds to be used to identify information obligations that should be excluded from the analysis. The consultation undertaken with some member states highlighted that the only possible option would be to set thresholds expressed in number of hours, in line with the Danish SCM variant31. However, the threshold set by the Danish government, i.e. laws imposing less than 1000 hours of administrative work per year, was found to deserve further scrutiny, as its 30 Operational manual, passim. 31 See below, section 2.2. – 22 – application at EU level might create concerns as regards the accuracy of assessing administrative burdens imposed on SMEs. This is even more important, as the EU model constantly refers to SMEs as the category most affected by administrative burdens. The operational manual of the Commission (March 2006) provides that “for administrative obligations requiring little equipment, if the amount of time per action is small and the frequency low, the obligation does not need to be quantified.” § In order to avoid ‘spurious accuracy’, Annex 10 to the Commission’s Impact Assessment guidelines mandates that DGs estimate a range (worst-best scenario) of administrative burdens, instead of a set figure. This arrangement might, of course, clash with the need to reach set figures when assessing cumulative burdens. In this case, the Commission has clarified that impact assessments and ex post evaluations may continue to refer to range of estimates when presenting their findings on administrative costs, but DGs should use the median figure when summarising findings. In this respect, no significant difference with national models can be said to exist in the reporting phase. § The proposed common methodology includes an identification of the main steps to be followed in the assessment exercise, which comprises 11 steps – as opposed to the 15 steps foreseen by the prototype SCM illustrated above, at Table 3. The eleven steps are summarised in Table 5 below. Table 5 – The phases in the Commission operational manual for ex ante assessment of administrative costs Phase 1. Preparatory analysis Step 1 Identification and classification of information obligations. Step 2 Identification of required actions Step 3 Classification by regulatory origin Step 4 Segmentation Step 5 Identification of the frequency of required actions Step 6 The relevant cost parameters are identified. Step 7 Choice of data sources and, where necessary, development of data capture tools Phase 2. data capture and standardisation Step 8 Assessment of the number of entities concerned (population) Step 9 Assessment of the performance of a “normally efficient entity” Phase 3. Calculation and reporting Step 10 The validated data are extrapolated to EU level. – 23 – Step 11 The results are presented in a report and the data set is entered in the overall database. Source: Operational Manual, Annex 10 to EC Impact Assessment Guidelines, 2005, p.4 The reduction in the number of steps is understandable, as the operational manual is, on one hand, targeting specific actors and, on the other hand, part of an overall scheme setting general rules and structures for support and quality control. The manual targets desk officers and units responsible for assessing the impact of a particular proposal. Therefore it does not cover aspects linked to ex post evaluation, baseline measurement of cumulative burden or reduction targets32. Moreover, the manual is but an annex to the general guidelines for Impact Assessments. The guidelines set general rules for data collection (how to build a questionnaire, interviews, expert groups, etc.). Procedures and structures in place (Roadmaps, Interservices Steering Groups, etc.) cover all IA aspects. There is not point repeating all this in a thematic annex. . However, there are a – limited - number of substantive differences in the analytical sequence proposed by the SCM phases and those in the Commission operational manual. The identification of the population is part of data capture and consequently is not listed as a preparatory step in the operational manual (see Step 8). The operational manual considers that a normally efficient entity can only be determined on the basis of data collected among other things through interviews. Insofar as standardisation is in turn based on the normally efficient entity, Commission Step 9 'normally efficient entity' subsumes the SCM Step 11 'interviews' and Step 12 'standardisation'. The description of SCM Step 10 'typical business' seems to be no more than an extension of SCM step 4 'segmentation'. Some of the activities that appear similar in the two models also hide differences in complexity: for example, the identification of the population and the segmentation are expected to be way more difficult when applied to the EU25, and the same can be said for the extrapolation of data to the EU level. Table 6 below compares the two models. 32 For instance, demarcation is by definition not a problem when assessing the burden of a single proposal. – 24 – Table 6 – Prototype SCM v. Commission operational manual Prototype SCM EU Model Phase 0. Start-up Phase 1. Preparatory analysis Phase 1. Preparatory analysis Step 1 Information obligations and administrative activities . Step 1 Information obligations. Step 2 Related regulation and demarcation Step 2 Identification of required actions Step 3 Classification of regulatory origi n Step 3 Classification by regulatory origin Step 4 Segmentation Step 5 Population, rate and frequency Step 4 Segmentation Step 6 Scope and data sources Step 5 Frequency of required actions Step 7 Relevant cost parameters Step 6 Relevant cost parameters Step 8 Interview guide Step 9 Expert review of steps 1-8 Step 7 Choice of data sources Phase 2. Analysis of time consumption and costs Phase 2. data capture and standardisation Step 10 Typical businesses Step 8 Population Step 11 Interviews Step 12 Standardisation Step 9 “normally efficient entity ” Step 13 Expert review of steps 10 and 12 Phase 3. Calculation and reporting Phase 3. Calculation and presentation Step 10 Extrapolation to EU level Step 14 Extrapolation to national level Step 15 Reporting and database entry Step 11 Reporting and database entry Finally, the classification of information obligations by regulatory origin is carried out differently in the two models. In the EU model, the proposed decision tree for the classification of regulatory origin was simplified. As a result, the steps followed to classify the regulatory origin of selected data requirements are significantly different from those reported above, in Figure 3. Instead of referring to A, B or C categories and their sub-categories33, the Commission distinguishes between four types of regulatory origins 1) IO/DRs arising from international (non-EU) legislation; 2) IO/DRs generated by EU legislation; 3) IO/DRs imposed by national legislation; and 4) IO/DRs arising from regional pieces of legislation. This solution was seen as more accurate (as it assigns responsibility to each level of authority without ambiguity), simpler (as it works with less categories and does not require to go through a complex decision tree) and more ‘pedagogical’ (as it relies on descriptive labels rather than letters). In particular, the distinction between EU directives and EU regulations was discarded because it proved rather unreliable – some directives detail which information must be provided and how, while some regulations leave the formulation of the specific data requirements to the discretion of Member States (e.g. the Intrastat regulation). Table 7 below reports the current EU standard reporting sheet, which highlights the four categories in which the IO/DRs have to be classified. 33Category A-EU directive, category A-EU regulation, category A-international + category A-EU directive, category A- EU regulation, category A-international – International SCM manual at 12. – 25 – Table 7 – EU standard reporting sheet Source: European Commission, Staff Working Document, SEC(2005)175, 16 March 2005, pp. 17-19. The observation of such differences is a suitable starting point to assess the feasibility of implementing the EU SCM without requiring major changes in the methodology currently followed by member states that rely on some variant of the prototype SCM. These and other problems will be analysed in more detail below, in Sections 3 and 4. 1.7.1 A comparison between the prototype SCM and the EU SCM Already in the March 2005 “Outline of a possible EU Net Administrative Cost Model”, the European Commission has presented a summary of the main differences between the Dutch SCM (on which, see also Section 2.1 below) and the proposed EU methodology. Some of these differences – for example, the assessment of burdens on citizens and the possibility of using the model for macroeconomic purposes, such as cumulative assessment – are today less evident than they were at that time, or do not exist anymore. But for the remaining aspects highlighted by the Commission, substantial divergence still exists, which can be partly explained by the different use of the two models, and the need to adapt the SCM to the peculiar needs and the greater complexity of applying the model at the EU level. Table 8 below summarises such differences. – 26 – Table 8 – Comparison between the Dutch SCM and the EU SCM Dutch SCM EU SCM Assessment of the costs of administrative Assessment of the costs of administrative Aims/purposes obligations imposed on enterprises and obligations imposed on enterprises, the citizens, distinguishing between national and voluntary sector, public authorities and non-national origins (ABC classification) citizens, distinguishing between international, EU, national and regional (i.e. subnational) origins Both microeconomic and macroeconomic use Both microeconomic and macroeconomic use “Costs imposed on businesses, when “costs incurred by enterprises, the voluntary Definition of complying with information obligations sector, public authorities and citizens in administrative cost stemming from government regulation.” meeting legal obligations to provide information on their action or production, “AB Citizen comprises the costs incurred by either to public authorities or to private citizens in complying with information parties.” obligations ensuing from government regulations. It includes both compliance with obligations and the exercise of rights.” One-off costs not taken into account in the One off costs may be taken into account measurement, but described qualitatively in the reporting Measurement includes ‘business as usual costs’ Measurement includes ‘business as usual costs’ but focuses on administrative costs instead of but focuses on administrative costs instead of burdens burdens Σ=Ρ•Q Σ=Ρ•Q Core equation Price (P) = tariff x time Price (P) = tariff x time Quantity (Q) = n. of businesses x frequency Quantity (Q) = n. of businesses x frequency Focus on labour costs Where appropriate, types of costs other than wages and overheads will be taken into account The SCM is applied to all regulatory proposals Ex ante application only to proposals imposing Scope and and acts in force. major administrative requirements and/or acts frequency identified as particularly burdensome by end users Applied to all administrative actions imposed Use of thresholds to identify most onerous by a piece of legislation, except for exceptional actions (and undefined) marginal costs Periodic review (4-5 years) Review and timeline for ex ante assessment defined on a case-by-case basis Periodicity of cumulative/sectoral ex post assessments yet to be defined. – 27 – High level of accuracy sought mainly through Use of range for administrative costs in ex ante Expected level of fieldwork and simulation impact assessments to avoid spurious accuracy, accuracy and data with median figures used in summarising results sources Use of national registers and statistics Application of the proportionality principle to establish optimal accuracy levels. Use of a sample of member states and extrapolation to EU level Estimates base on EU statistics, standardised ratios and data provided by the sampled member states. Decision tree in 7 steps Decision tree in 4 steps Classification of regulatory origin Each public entity assesses its regulatory Application of the subsidiarity principle: Division of proposals and sectoral legislation (with Commission assesses ‘upper bound costs’ and responsibility involvement of consultants) under the extrapolate national data to EU level supervision of the IPAL and monitoring by Actal (The sample of) member states must provide data on their implementation of EU legislation. Need for interoperable databases Need to streamline communication between Commission services and responsible national authorities. Assumptions are clarified at the outset. Inclusion of caveats clearly drawing attention Methodological to the underlying assumptions and their effect caveats on the accuracy of the assessment. 3-5 dedicated and approx €300,000 per ministry 14-40 hours of work over 4-24 weeks per ex ante Workload and cost assessment Around 60 people involved in public administrations (IPAL, Actal, ministries) 1,600 hours/week per year for the Commission’s central policy unit(s) In 2002, overall cost for the baseline measurement was approximately €3mn (done by consultants) As emerges from Table 8 above, the feasibility of developing an EU common methodology for assessing administrative costs still depends on whether an efficient and effective solution can be found to pending methodological, procedural and organisational issues. The Commission identified areas for optimisation of the model in its report on the pilot phase issued on October 21, 2005. These issues are presented in the next section. 1.7.2 Pending issues after the pilot phase As already mentioned in the previous sections, developing a SCM to be applied both ex ante and ex post at EU level is a challenging task by all means. The EU SCM is now operational and seems promising in many respects, but it is particularly important to – 28 – make swift progress on priorities for optimisation. Such areas are currently identified by the Commission as follows. 1. Database interoperability: to ensure the interoperability of national databases on administrative burden and access for the Commission. 2. Country distribution: identification of weighting systems for assessing EU-wide costs on the basis of limited national data (e.g. country distribution). 3. Accuracy: identification of the average margin of error of administrative cost assessments. 4. Standard ratios: identification of standard ratios for overheads, training costs and learning curves and for costs corresponding to normal business operation, among other things. 5. Thresholds: identification of specific threshold(s) below which quantification is not necessary (minimum thresholds for the application of the model). 6. Extension to citizens: possible adjustments of the model when assessing administrative costs put on citizens. 7. Guidance on borderline cases: possible difficulties to distinguish information obligations from the other regulatory costs and how to overcome them. 8. Standardisation of IOs and target groups: looking at possible shortcomings of the typologies of IOs and required actions used in the EU operational manual; examining the need for a typology of target groups. 9. Exchange of data: Organising optimal exchange of data between the Member States (including their regional authorities) and the Commission. 10. Target-setting: the issue of target-setting still needs to be addressed and agreed upon. Whether an overall target can be set at pan-European level for the reduction of administrative burdens imposed by legislation within a given timeframe is one of the issues that will have to be explored in the near future. This Report will offer suggestions on these and other pending issues in Section 4 below. – 29 – 2 COUNTRY CASE STUDIES In this section, we present the findings of four case studies carried out during the months of August-October 2006. We analysed the main features of the model adopted by the four member states that have completed so far the baseline measurement – i.e. The Netherlands, Denmark, UK and the Czech Republic – in order to identify similarities and differences between the variants of the SCM adopted, highlight the potential for cross-country comparison, comment on the possible co-existence of these models with an EU common methodology, and report the results reached in the national measurement with specific emphasis on burdensome EU legislation. 2.1 The Netherlands Since 2000 onwards, the Dutch government has intensified its efforts to reduce the administrative burdens on businesses and citizens and, according to the advice of the high level Slechte Committee, it decided to quantify burdens on businesses. The government also set up the Dutch independent Advisory Board on Administrative Burdens (ACTAL) by government decree (on May the 1st 2000). Notwithstanding the efforts made by the government, the overall reduction was still modest: in 2000 the total red tape reduction was estimated to be around €1 billion, i.e. 6-7% compared to the 1994 level. Meanwhile, the administrative burdens were increasing as a result of new legislation; in particular, the economic situation became more difficult for private businesses, which asked the government to reduce burdensome obligations stemming from the regulation. Accordingly, in 2003 the Dutch government re-launched the regulatory reform as one of its top political priorities. It committed itself, in the coalition agreement that was the basis of the second Balkenende Cabinet, to reduce red tape by a quarter compared to the baseline measurement by the end of its term (May 2003-May 2007). This meant reducing more than €4 billion of the €16.4 billion total administrative burdens estimated in the Netherlands (at 1 January 2003). The task to coordinate the needed activities was laid in the hands of the Ministry of Finance in close collaboration with the Ministry of Economic Affairs. The Interministerial Project Unit for Administrative Burdens (IPAL) was placed under the authority of the Minister of Finance. Since then, progress of the reduction project is monitored via the budget cycle: the annual Budget Memorandum and the Ministerial budgets (September), and the annual reports (May). This operation requires the involvement and support of the entire Cabinet. All Ministers support this approach and take responsibility for the realisation of their own plans. Currently the Dutch government is preparing its final letter about the state of play of the administrative burden operation for Parliament. The project is on schedule. About half of the reductions are already implemented, whereas the rest will follow in 2006 and 2007. Originally the final letter was planned for spring 2007, at the end of the Cabinet. However, the Dutch Cabinet resigned this summer, and elections will take place in November 2006. The fight against red tape is expected to be a major issue also in the new government’s agenda. – 30 – In 2003 the Cabinet developed a methodology to quantify administrative burdens – called the Standard Cost Model – which is among the earliest and most thoroughly applied systems to measure administrative burdens in European countries.34 In the Netherlands this quantitative methodology has been developed to measure administrative burdens stemming from both European/international and national legislation by identifying the information obligations imposed by legislation.35 Burdens are quantified in time spent to fulfil the requirements as well as in monetary terms and are considered as costs imposed on businesses when complying with information obligations stemming from the government regulation. The Standard Cost Model works in three steps: 1) an in-depth analysis of the data transfer between a business and the authority which are isolated and defined; 2) the determination of the time involved in each “data transfer” and the level of the person performing it; 3) the computation of the data to produce cost estimates.36 The Standard Cost Model was conceived for both ex ante and ex post applications. 2.1.1 The Standard Cost Model According to a step by step approach, the guide for classifying administrative burdens according to national (and international) origin states the various steps to classify information obligations (IOs): • preparing the survey to obtain an overview which indicates the legislative domains, the missing information required for classification in categories and an inventory of the policy officers and jurists responsible; • clarifying the survey to organise the knowledge transfer for classification of the legislation and regulations in categories; • classifying the information obligations into three categories according to the ‘decision tree’. Each ministry has classified the information obligations into three categories according to the ABC categorisation to determine the origins of administrative burdens37; • consolidating results of the classifications and analyse the focal points to summarise the results of classifying legislation. According to the Dutch SCM manual (2003), the Netherlands have some standardised information obligations which are selected on the basis of the measurement 34 From Red Tape to Smart Tape Administrative Simplification in OECD Countries, OECD, Paris, 2003, p. 44 35An information obligation is the passage in a law or regulation text that requires a business to provide or draw up information, and make this available, i.e. a duty that the business cannot avoid without violating the law. 36 See From Red Tape to Smart Tape Administrative Simplification in OECD Countries, OECD Paris, 2003, p.45 37All Ministries have established the origins of information obligations (national /international). To support the Ministries in classifying the administrative burdens according to their origin, the Dutch Inter-Ministerial Legislative Burden Department (IPAL) has developed a guide. This guide contains a ‘decision tree’ with whom the information obligations underpinning the administrative burdens can be broken down into three categories. See: Dutch Legislative Burden Department (IPAL): Administrative burdens in an (inter)national perspective Guide for classifying administrative burdens according to national (and international) origins, The Hague, August 2003. – 31 – experience with the help of consultants who carried out the measures: however, the list of IOs is not exhaustive. 2.1.2 Baseline measurement As already mentioned in Section 1, a baseline measurement is an ex post measurement of the overall administrative costs that enterprises experience at a given point in time in following a current set of rules. The Dutch program, including the baseline measurement, officially started in 2003 and initially involved only private businesses and some semi-private businesses. Furthermore, the Dutch Cabinet launched a pilot project to tackle the intergovernmental costs among public authorities and to measure burdens on the voluntary sector. Finally, as will be clarified in more detail below, the Dutch government has launched a measurement of the administrative burdens imposed by legislation on citizens. As a result, if compared with the EU common methodology, the Dutch model has a very similar scope, but does not cover public administrations.38 The SCM is suitable for both a limited and a full scale measurement covering all legislation generating administrative costs.39 In particular, the Dutch approach provides for a full scale baseline measurement for private businesses. To the contrary, in measuring administrative burdens for citizens a Pareto distribution is assumed (20% of the regulation causes 80% of the administrative burdens).40 The selection of the burdensome legislation was made on the basis of a measurement of the number of contact moments between citizens and the government. A baseline measurement report is drafted per ministry. Each department measured the administrative burdens which fall under its responsibility. In borderline cases, where the same regulatory requirement can be attributed to more than one ministry or department, generally the Dutch approach prefers to split it up, fifty-fifty. If this is not satisfactory, departments have to reach an alternative agreement on a case-by-base basis. The administrative costs are noted down per law. Each law is sometimes split up in consideration of the sectors and the sizes of companies affected which determine the needed segmentation41. A business panel selected by the consultants is used to select businesses to be included in the samples which are made per type of legislation under consideration. In case the needed type of company is not present in the panel, the consultants choose for interviews companies that are considered to be relevant. 38 See above, section 1.7 for a comparison between the Dutch model and the EU common methodology. 39The limited scale approach can be used to gain some technical knowledge and practical experience with the methodology, before deciding whether to proceed or not at a full scale level. 40 According to the Pareto principle, only the most burdensome legislation is measured and extrapolated. 41 Segmentation is required when companies comply differently. – 32 – 2.1.3 Methodology 18.104.22.168 Scope of measurement As already mentioned, the baseline measurement covered businesses. In the SCM Manual a definition is given of what businesses are: “The term businesses should be taken to mean every sector of the economy, with the exception of the public sections of public administration, government services and compulsory social insurances and education. However, the private sections of the above–mentions exceptions do constitute businesses. The basic criterion is the question of whether the organisation covers its own costs. If it does, it is a business” [The Standard Cost Model, 2004, p. 10] To implement this definition sector codes from the national account of Statistics Netherlands (In Dutch: Centraal Bureau voor de Statistiek, CBS) have been used. Healthcare organisations and educational institutes can be subject to the baseline measurement as long as these organisations fulfil the definition of businesses. This definition implies that for instance compulsory education is excluded, but private educational institutes are included. Nevertheless the areas of criminal law, subsidies and public procurement and defence are not covered by the measurement. Costs stemming by inspection are measured, but not systematically, unlike what happens in the UK42; the measurement in this area regards only the costs formulated in the legislation except for the provisions which the inspection authorities could formulate according to the legislation. Moreover, criminal law does not fall in the context of administrative cost exercise because this field is strictly linked to the legal system. Finally, the subsidies and public procurement is included later on and the defence area is considered not relevant for the measurement. Apart from these areas, no other areas were in principle excluded from the measurement: the Dutch measurement included all information obligations that result from generally binding regulation. Table 9 shows the results of the baseline measurement by ministry at 1 January 2003, and the ABC classification of the administrative burdens for each policy domain (ministry). 42 See below, section 2.3. – 33 – Table 9 – Total Administrative Burden of the baseline measurement at 01-01- 2003, by ministry according to the ABC classification*) (€ million) Finally, the Dutch model is conceived to measure administrative costs, not just administrative burdens. This means that it measures also obligations that would be complied with even if the compulsory regulation would be removed. In addition, also voluntary obligations were measured. This broad approach was chosen to include all unnecessary red tape. 22.214.171.124 Costs and tariffs43 Administrative burdens are part of the total compliance costs. But not all compliance costs are regarded as administrative burdens. The definition includes only a number of categories of costs that are sufficiently concrete and objective/measurable to be useful in terms of quantifying administrative burdens. When examining the size of administrative burdens, the matter of whether a business views specific information obligations as a burden in the subjective sense of being a nuisance ('irritation costs’) is not important. What needs to be ascertained is whether the burden is an objective burden in the sense of a ‘cost item’. Possible benefits of a regulation are not explicitly considered in the measurement itself, although the relative merit of burdensome regulation is taken into account when formulating cost reduction plans. Also lost turnovers due to an information obligation are not included in the administrative burdens. Substantial one-off costs that a business must invest in order to comply with an IO are quantified and spread over the depreciation period. Negligible one-off costs need not be quantified, but it should be possible to obtain at least qualitative insights on them. Costs resulting from trying to keep up with the frequently changing existing legislation and regulations are considered to be administrative burdens. As regards ex ante measurements, it is important to consider the costs to a business of keeping up with new legislation and regulations. These costs are an administrative burden, but do not need to be quantified in advance. 43 The Standard Cost Model A framework for defining and quantifying administrative burdens for businesses, International working group on Administrative Burdens. August 2004: p. 14-16. – 34 – Administrative costs are estimated on the assumption that legislation is complied with for 100%, despite the fact that in practice, not all businesses comply (fully) with all of the legislation. However, when there is sufficient evidence that not all businesses comply with all of the legislation, the 100% compliance assumption is not applied. A number of businesses register data for more purposes than just the obligations arising from legislation and regulations. The total costs of multi-purpose registrations should be allocated to the legislation or regulation concerned as an administrative burden. As far as demarcation issues are concerned, if there is an overlap between two areas of legislation (the same information is used by two different parties), the costs of the reporting are shared on a 50%-50% basis between the two legislative areas, unless the involved parties arrive at another solution on the basis of measurements. In calculating the administrative burden, the costs of an individual administrative action (P) are determined by multiplying the time to fulfil this administrative action with a tariff. There are two types of tariffs: internal and external. The internal tariff is the hourly rate of the person(s) in the business who deals with the IO. The external tariff is the hourly rate of the person(s) outside the organisation who deal with the information obligation. This can include contracting out and the hourly rate of the accountant, for example. These costs are usually offset at a commercial rate on the basis of wage costs. The internal tariff is made up of the following components: • Gross wage + Wage costs • Material and overhead costs Material costs are taken to mean the costs of all materials purchased in order to satisfy the administrative obligations. This can include the costs of accommodation, materials needed for a registration system, detailed drawings and the like. Overhead costs are all the costs associated with the use of office materials, depreciation of desks, computers, facilities for ancillary departments, etc. In general and where relevant, these costs are estimated using a mark-up percentage on the internal tariff of the gross wage costs. In the Netherlands the overhead tariff is not formally standardised. At the start of the measurement an overhead percentage of 25% of the gross wage was used. Now sometimes higher percentages are used. When no information is available to calculate a tariff in an ex ante exercise, it is possible to make use of standardised tariffs. These standardised internal tariffs are as follows:44 low level job € 30/hour middle level job € 45/hour high level job € 60/hour 44 Meten is weten, p. 31 – 35 – The external standardised tariff is € 150/hour. These tariffs include wage costs and material and overhead costs. 126.96.36.199 Procedural Issues Central in the Dutch organisational structure are the Minister of Finance and the IPAL. The latter was in charge of coordinating the measurement process and developing a common methodology, i.e. the Standard Cost Model Manual.45 This manual needs to be applied by all the ministries in measuring the administrative burdens and measuring the effects of the reduction policies. The authority of IPAL as the coordinating department rests on the position of the Minister of Finance. In the Dutch coalition cabinets the Minister of Finance has always held a central and powerful position. Strong political input was provided by the cabinet Balkenende, under which the issue of reducing administrative burdens ranked very high in the political agenda. On the other hand, the measurement and reduction of administrative burdens for citizens is coordinated by the Minister of Government Reform and Kingdom Relations. This ministry developed the methodology for measuring AB on citizens,46 and safeguards its uniform application amongst ministries. The ‘citizens manual’ shares most of the starting points with the Standard Cost Model Manual; however, the administrative burdens for citizens are expressed only in hours, with no segmentation based on occupational groups, and this results in the application of a uniform cost per hour when transforming time-based measures into actual administrative costs. 188.8.131.52 Identifying Information obligations No compulsory standardised list of information obligations is defined in the SCM Manual. In the process of measuring AB some standardization took place. Information obligations were defined on the basis of the measurement experience that IPAL gathered before 2003. Consultants assisted in carrying out the measurement. To classify information obligations as EU and non EU related, the ABC methodology was applied. This methodology was laid down in a separate manual.47 When an information obligation can be contributed to more than one ministry or department the obligation was divided amongst the ministries or departments in equal terms. When this was not satisfying the ministries can make other arrangements by themselves. The baseline measurement was carried and reported by each ministry, because the ministers are responsible for the reduction policy of their own ministry. Per ministry the administrative burden is recorded per law. An information obligation sometimes 45 Quote 46 Quote 47 (Reference) – 36 – must be split up to be able to diversify the administrative burden according to sector or to firm size. It was recognised in the baseline measurement that segmentation is needed whenever the same IO can exert a different impact on different sectors and different businesses. This classification of administrative burdens by ministry is used in reporting to the Parliament and other stakeholders. But in communication with important stakeholders like industry associations, the sector or target group was used as a unit of reporting, because this reflects the business perspective better than a ministry-wide view. 184.108.40.206 Selecting businesses The businesses were selected on the basis of information from business panels organised by the consultants. Samples of types of businesses were made per type of legislation under surveillance. If a type of business was not present in the panel, the consultants directly approached businesses belonging to this type for interviews. 220.127.116.11 Data sources and collection methods Different data sources from the selected business were used. The Dutch zero-based measurement is a combination of several instruments for collecting data. These instruments are: desk research, discussions with representatives of relevant business sectors and field research. Different phases can be distinguished in the establishing the information obligation: • Desk research was used to find the IOs that arise from a given regulation. The purpose was to reach a preliminary judgement on how to separate IOs from ordinary business administration and making a first proposal for the blue-print of administrative activities. • Discussion with representatives of relevant business sectors were organised to verify and create support for the list of DRs (these are units of administrative activities necessary to fulfil an IO) and some expert interviews to check the blue-print of administrative activities; • Field research (face-to-face interviews) were used to collect cost parameters, like time parameters. These face-to-face interviews were addressed mostly at bookkeepers, controllers and accountants; • Desk research was undertaken for collecting parameters to extrapolate the results of the measurement to a national level. These parameters are, amongst others, the number of entities (enterprises, employees, transactions, etc.) in the relevant field and the enforcing institutes like tax authorities, social insurance agencies, bureaus of statistics, chambers of commerce. • The “normal face-to-face interviews” and/or the “stop watch method” refer to the field research to collect time parameters only. During the experimental phase (1992- 1994) of the SCM, then called Mistral®, several methods of collecting time parameters bottom-up were tested, e.g. multi-moment, registration by the – 37 – respondents themselves, stopwatch and time registration by computer.48 At the same time several top-down instruments to collect time parameters were tested, such as mail questionnaires, structured face-to-face interviews by trained and qualified researchers and structured face-to-face interviews by external (outsourcing) bookkeepers and accountants. These instruments were evaluated/tested on three criteria: accurateness of the time measurement, the content of the results (more or less distortion by perception, possibilities to detail) and cost-effectiveness. In 1994 these instruments were tested in a laboratorial setting involving some 40 respondents from affected businesses. The major result of these laboratorial experiment was that the results of the bottom-up approach were more accurate then those of the top-down approach. In detail: • the variation of time parameters resulting from a top-down approach were larger then in case of a bottom-up approach; • After a correction for outliers (4 highest and 4 lowest results) the time measurements of the bottom-up approach appeared much more stable then those resulting from the top-down approach; • The coefficients of variation of the bottom-up approach were smaller then those of the top-down approach; • After correction for outliers, the variation coefficients resulting from the bottom-up approach were much more stable then those resulting from the top-down approach; • The bottom-up approach of time parameters after the bottom-up approach showed a better fit to the Lorenz curve then was the case for the top-down approach. During face-to-face interviews most respondents think in categories of one minute or more. As a result, the stopwatch is especially recommended when the numbers of entities are in the order of millions or higher and the administrative tasks to be measured are expected to exhibit a high frequency and to take less than 1 minute to accomplish. However, in the Netherlands this happened only three times in a period of 10 years. In almost all cases the face-to-face interview appears to be good enough for accurate time measurements. Apart from face-to-face interviews, the main data collection methods used were telephone interviews, focus groups, expert assessments and business panels. Other data collection methods were used only occasionally - including focus groups and consultancy studies. 48 EIM Business Policy Research was involved in the experimental phase – 38 – 18.104.22.168 The ‘normally efficient business’ A goal of the government is to measure administrative burdens that occur because of legal prescriptions, not because of inefficient operations taking place within a given firm. Accordingly, when measuring the time needed to perform a given activity, the performance of a normally efficient firm is taken into account: this does not mean considering the most efficient firm, but a firm that can be regarded as representative for the population of firms under study. A frequently used approach is the determination of a “typical firm”. In this approach the law is the starting point. The standard costs businesses face to comply with the IOs laid down in the law are calculated. These are not to be taken as costs from a business economics point of view – e.g., opportunity costs or inefficiency costs. The SCM does not reckon benefits at firm level as a result of complying with the obligation (e.g. lower tax payments or subsidies) and marginal costs: it only quantifies the costs that would disappear if the legal prescription was repealed. As such, the model employs a public administration point of view: on the basis of the relevant information requirements, the time spent by a typical firm in complying with the obligation is measured. The concept of typical firm deserves some further explanation. From the business economics perspective there are no typical firms. The construction of this artefact is however necessary in order to measure administrative burden from the public administration perspective. In case administrative burdens of a specific law have to be measured, the respondents (i.e. businesses or their representatives) are selected. The Dutch SCM does not use random samples to select respondents for time measurements. The selection of typical firms depends primarily on the legal specifications of the law under surveillance. The total population of enterprises affected by a law is segmented into a limited number of groups of typical firms, according to the legal specifications in that law. During the selection process of enterprises for in-depth interviews, these different segments are taken into account. All different (but more or less common) types are selected in a substantial rather than a representative manner. Corrections for over- or under-representativeness are made at a later stage (re-weighing with the correct population figures). The experts validate the rationale behind the selection of typical firms. Consequently, in-depth interviews are held with a number of enterprises per group. The determination of typical firms can be illustrated by an example. The Annual Account Act obliges enterprises to prepare an Annual Account. The type of account (simple or long) enterprises have to prepare depends on enterprise characteristics as defined in the Act: number of employees and total revenues. The number of enterprises to be interviewed was selected as follows. First, desk research showed that 40% of the enterprises have the obligation to file a simple account sheet, whereas the remaining 60% has to file a long sheet. Secondly, for measuring the compliance cost it was important to estimate the percentage of enterprises that hired an external party to prepare the annual account; it emerged that 75% of enterprises outsource this – 39 – obligation to an accountant or an administration office. On the basis of these two criteria, four types of typical firms could be distinguished, and the sample for interviews could be constructed as follows (Table 10 presents the number of enterprises per category). Table 10 - Example – selection of sample firms In this example, 15 interviews would provide information regarding enterprises that outsource the obligation, 12 interviews would provide information regarding the “long” account, etc. Businesses were segmented when necessary according to firm size, industry sector and the outsourcing/in-house provision of the requirement. Segmentation is only necessary when the way to comply with the IO is assumed to differ in the various segments. 22.214.171.124 Standardisation of measurements In the Dutch baseline measurement of administrative burdens on businesses, no standardisation of time per data requirement of an information obligation took place. In the measurement of the administrative burdens on citizens the time per activity has been standardised. The tariff attached to an hour spent was not standardised: however, in the next baseline measurement to be carried out in 2007, 3 rates will be applied for the internal tariff: a high tariff, a middle tariff and a low tariff.49 126.96.36.199 Who does what In the Netherlands the Ministries are responsible for their own baseline measurement, i.e. each Ministry organises its own baseline measurement with a consultant of his own choice. All ministries are obligated to follow the SCM Manual. A board is always involved in the process of a measurement and IPAL will always be represented. Generally, the parties involved in most of the steps of the measurement 49The internal tariff is the hourly rate of the person(s) in the business who deal with the IO. The external tariff is an hourly rate of the person(s) outside the organisation for which the AB is calculated. See SCM Manual, 2004, p. 19. – 40 – process are: the responsible ministry, the coordinating unit, the consultant and businesses (and/or business organization). See also Table 11 below. Table 11 – Who does what in the 15 steps of the SCM, The Netherlands Consultant(s) Coordinating Ministry(ies) businesses Central Other/ Phase Step Unit Description Identification of business-related regulation to be Phase 0 0 X X included in the analysis Identification of IOs/DRs/administrative activities and 1 X X classification by origin 2 Identification and demarcation of related regulation X X 3 Classification of information obligations by type X X 4 Identification of relevant business segments X X X Phase 1 5 Identification of population, rate and frequency X X 6 Business interviews v. expert assessment X X X 7 Identification of relevant cost parameters X X X 8 Preparation of interview guide 9 Expert review of steps 1-8 X X X X 10 Selection of normally efficient business X X X 11 Business interviews X X Phase 2 Completion and standardisation for each segment by 12 X X X activity 13 Expert review of steps 10-12 X X X X 14 Extrapolation of validated data to national level X X X X Phase 3 15 Reporting and transfer to database X X 188.8.131.52 External consultants and budget More than one consultant was involved (4-5). These were experienced consultants in the field, which were selected by a tender procedure. Each department could select the consultant of its choice. All consultants used the SCM Manual drafted by IPAL. The full baseline measurement took about 6 months, although some areas were already measured when the operation commenced. As regards public administrations, two central departments are mostly involved in the process, ACTAL and IPAL. The actual measurement of the AB is decentralised. Normally 3-5 persons per ministry are involved. Furthermore, different persons per ministry work on the same measurement. So the costs per measurement differ from Ministry to Ministry. The actual figure depends on several factors: – 41 – • Amount of legislation. Ministries differ in size and there are also differences in the amount of legislation they are responsible for. A ministry that has few rules with administrative burden has relatively low measurement costs. • Complexity of legislation. The complexity of the legislation affects the way data has to be collected. More complex legislation can require more desk research and can require more interviews in order to get a representative outcome. • Target group of legislation. The target group of legislation can for example be one or several sectors of industry or certain groups of enterprises (SMEs). In the context of measuring administrative burden this can imply that the groups carry out their information obligations differently. For example, situations in which small firms fulfil an obligation much more efficiently than large firms. If this is a substantial difference this is an aspect that should be taken into account to get a representative outcome that reflects typical firms in practice. In general in can be said that the costs per legislative area in the Netherlands varied from €100,000 to €300,000. The total cost of measurement of administrative burden for Dutch enterprises was approximately € 2.5–3 million euros for all nine participating ministries together. These are one-off costs, not annual costs. After finishing the baseline measurements, only changes (due to new legislation and to reduction plans) have to be measured. 184.108.40.206 Administrative burdens for citizens50 In its Outline Agreement, the former Cabinet has proposed that the administrative burden for citizens should be reduced. This Cabinet objective forms part of the 'Modernizing Government’ action programme51 as formulated by the Minister of the Interior and Kingdom Relations and the Minister for Government Reform and Kingdom Relations. The target is that by 2007 the administrative burden on citizens (AB Citizen) should have been reduced by 25% compared to the base line measurement with the reference date 31 December 2002.52 The Ministry of the Interior and Kingdom Relations is responsible for coordinating the activities carried out by the various ministries to achieve the AB Citizen target.53 The cooperation between ministries has been given tangible form in the interdepartmental project group ICRAL, which includes representatives from all ministries whose regulations give rise to ABs Citizen. Additionally The Advisory Board on Administrative Burdens (ACTAL) has been given the task of advising on the administrative burden for citizens. 50This section is based on “Standard Cost Model. Administrative burden for citizens.” Ministry of the Interior and Kingdom Relations. September 2004. 51 Parliamentary publication: TK, 2003 – 2004, 29362, n. 17. 52 As stated in the memorandum from the Ministry of BZK to the House of Representatives. 53As adopted by Cabinet in the Administrative Burden for Citizens Reduction Action Plan (Plan van Aanpak Administratieve Lastenreductie Burger). – 42 – Administrative Burden for Citizens are defined as those costs that are incurred by citizens in complying with information obligations ensuing from government regulations. This includes both compliance with obligations and exercise of rights. Costs that do not qualify as AB Citizen are: • The costs that can arise owing to the, possibly poor, quality of the service delivered by the government. For instance: long turnaround times and poor accessibility • The perception of citizens, for example that the government is delivering an inefficient service or has too little focus on the client. • The costs incurred by the government in dealing with requests or objections and gathering and processing information supplied by citizens. • The non-compulsory contacts between citizens and government, for example contacts not based on a statutory information obligation. AB Citizen relate to the time spent, expressed in hours, and to out-of-pocket expenses, in euros, incurred by the citizen. Measurement of the AB citizen is based on identification of actions carried out by citizens to comply with information obligations originating from a certain regulation. The time spent performing these actions and the costs incurred in doing so together constitute the AB Citizen of the IO in question. An IO is a statutory duty to supply information. Supplying information includes the gathering, processing, recording, keeping and making available of information. Citizens can be individual citizens and organised citizens. Individual citizens are persons of all ages. Organised citizens are not-for-profit organisations and associations whose funds largely come from voluntary contributions (gifts and donations) from households and income from assets. As a result, the Dutch measurement of AB citizens also covers non-profit organisations such as charities, which in other countries are measured together with administrative burdens for private (for profit) businesses. Each information obligation applies to one or more target groups. The Dutch Parliament has stipulated a number of target groups for which the AB Citizen must be quantified. These are: • General citizens This target group comprises citizens who are aged 18 or over. • The elderly These are citizens who were aged 65 or over. • Benefit claimants This target group consists of citizens who are entitled to benefit specific regulations. • Chronically ill and disabled persons. • Organised citizens This target group contains civil-society organisations. The measurements carried out by ministries and other government agencies must provide an insight into the AB for the above five target groups. Regulations are adopted by various government bodies. In the context of the definition of AB Citizen, the term ‘government’ is understood to include: community – 43 – organisations (EU), government ministries, provincial authorities, local authorities and water Authorities, Autonomous administrative authorities (In Dutch: ZBO). In order to obtain clarity as to whether the AB Citizen originates from EU/international or from national regulation also here the A/B/C classification is used. Furthermore, in the baseline measurement the total AB Citizen is distinguished based on the degree of policy discretion enjoyed by different levels of government. Four categories are distinguished: • Category I. Central government regulations. • Category II. Shared government without policy discretion. The regulations are drawn up directly by central government. The administrative authority has an exclusively implementing task and therefore no policy discretion • Category III. Shared government with policy discretion. The regulations are drawn up by central government, but the administrative authority has the freedom to set additional rules and is free in the method of implementation. • Category IV. Autonomous regulations. An administrative authority, which determines its own policy as well as the associated rules and information obligations, draws up the regulations. The AB Citizen from categories I and II is attributed to ministries. The AB Citizen from categories III and IV is attributed to administrative authorities.54 The measurement of AB Citizen is broadly comparable with the calculation used for AB Business. However differences exist. The main difference is that the AB Citizen is expressed in two dimensions: time (hours) and costs (euros). The total time and costs are determined by multiplying the time and costs (T and K) that have to be expended in complying with an information obligation for one citizen by the frequency of compliance (Q). Time spent complying with an information obligation includes: travel time to and from a government organisation, time spent completing name and address details, time spent looking for and entering a social security number, time spent asking for proof of registration with a municipality or time spent applying for forms via the Internet. The costs included in the AB Citizen are the money actually spent to comply with an IO. Opportunity costs (‘What I could otherwise have done with my time/money’) are by definition not included in AB Citizen. It is not important, when identifying the extent of the AB Citizen, whether a given information obligation is perceived as a burden by the citizen in the subjective sense of irritation. The AB must therefore be determined objectively as an average investment in time and costs for a citizen. Possible benefits experienced by citizens due 54Regulations implemented by autonomous administrative authorities are attributed to the ministries. Regulations implemented by other organisations are attributed to the shared government bodies. – 44 – to an information obligation are left out of consideration. This also applies for rights that are exercised. Getting familiar with legislation and regulations is also part of the compliance with an IO. For example, the time taken to learn about the fiscal rules for completing a tax return is seen as part of the time spent on completing the return and therefore as an administrative burden for citizens. With the aid of input from citizens, an average time investment is determined for this. Measuring AB Citizen is concerned with the overall costs of meeting the information obligations. It thus makes no difference whether the citizen also uses the information provided or the action carried out for their own purposes (mixed function). What is important is that the information obligations are laid down in legislation and regulations. Where there are two legislative domains requiring the same information from the citizen, the costs are attributed to each domain on a 50-50 basis, unless a different arrangement is agreed between the two domains. Time and cost parameters are determined mostly through citizen panels, possibly supplemented by interviews. When preparing and carrying out citizen panels, specific attention is devoted to the representativeness of the data gathered. After gathering the data from a citizen panel, the data is checked with the bodies responsible for implementing the regulation in question. This step in the procedure also enables possible reduction proposals to be identified. However, as often only limited data are available, the calculation of the AB Citizen ensuing from proposed regulations is generally based on estimates and forecasts by experts. When implementing the findings, therefore, allowance will always have to be made for the assumptions and principles used. 2.1.4 Targets An overall reduction target was set before starting the measurement exercise during the formation of the former Cabinet (Balkenende II). The political decision-making process to form the government after the elections of 2003 resulted in the choice for a 25% reduction target, which was considered politically feasible and taken up in the Coalition Agreement of the government. As a result, the target was mainly set through a political decision. The target was expressed as a percentage of the baseline measurement of the administrative burdens for businesses as of 31 December 2002. These were estimated to reach €16.4 billion, i.e. approximately 3.6% of Dutch GDP. The 25% reduction target was then 25% of €16.4, i.e. €4.1 billion. As a result, at the end of the Cabinet mandate in 2007 the administrative burdens for businesses should be € 4.1 billion lower compared to the level in 2002. – 45 – It must also be specified that the reduction target is a ‘net’ target: new administrative burdens imposed on businesses in the period 2003–2007 will be included in the total amount of administrative burdens to be reduced. Subsequently, the target was differentiated per Ministry: some ministries have a target of 15%; others have a target of 30%. But on average and overall the target is 25%. The projects were identified on the basis of information from ministerial committees, information from an ad hoc website (meldpunt AL), the project model companies, suggestions from the Parliament and information of a specific committee (Commissie Stevens).55 2.1.5 Results Table 12 below summarises the results of the reduction operation as they were reported by a letter of 14 April 2006 from the Cabinet to the Second Chamber of the Staten-Generaal.56 Net reductions are reported. At the moment the letter was sent to the Second Chamber, administrative burdens had been reduced by €1.9 billion. The reduction involved approximately 100 regulations. At that moment, there were still 70 regulations awaiting for measures to reduce the corresponding administrative costs. These further measures are expected to lead to a further net reduction of € 2.9 billion. (Pareto distribution!) Remarkable is that by the end of 2005 the cumulative net reduction of the Ministry of VWS was negative. The Cabinet concluded that in the budget year 2005 the reduction effort had not met the target. However, it is expected that in 2006 and 2007 the target will eventually be met. 55 [TK 29515, 135 bijlage De ondernemer Centraal] 56 [TK 2005-2006, 29515, 135] – 46 – Table 12 – Realised net reduction and prognoses by Ministry (€ million)*) These results are in line with one of the main conclusions of the report of the Netherlands Court of Audit.57 The Court of Audit analysed 24 government measures, which included 20 measures that were taken in 2003-2004 to reduce the administrative burden and 4 new regulations that resulted in increased administrative burden.58 For the latter, the Cabinet took compensating measures. The Court also studied the effects that the reduction policy has had so far in the businesses community. Importantly, the Court of Audit analysed also the methodology applied to measure administrative burdens, and concluded that the Cabinet has correctly applied the methodology set out in the SCM. The Cabinet is expected to comply with the 25% reduction target. One main criticisms expressed by the Court of Audit, however, is that the effects of the reduction policy on businesses are less evident than could be inferred from governmental reports. 2.1.6 Most burdensome EU legislation The results of the Dutch baseline measurement are presented below, in Table 13. We show the results in descending order according to burdens in Category A and, secondarily, in Category B obligations. As emerges from the table, in the Netherlands the areas (domains) most affected by EU/international legislation are health protection, private law and fiscal law. As international (non-EU) legislation is estimated by IPAL to account for less than 10% of total burdens in Category A, it can be fairly stated that administrative burdens in these sectors are mostly originated by 57 [TK 2005-2006, 30605, 2] 58Which included, for example, measures aimed at simplifying commuting regulations and at introducing digitial car registration certificates. – 47 – EU legislation. Other sectors that are significantly affected by Category A obligations are working conditions legislation, pricing laws, health care legislation, financial markets and agriculture. Also as regards Category B, IPAL estimates that non-EU legislation only accounts for a minor share of total burdens – i.e. less than 10%. Policy domains that are most affected by Category B obligations generated by EU legislation are fiscal law, private law, financial markets legislation, spatial planning and environmental licenses, agriculture and legislation on working conditions. – 48 – Table 13 – Overview of results by department, domain and origin (€ million)*) Source: IPAL – 49 – 2.1.7 Conclusion The Dutch model to reduce administrative burdens can be considered as a success story, especially since the political target set will be eventually met within the specified deadline. As already mentioned, this target is a 25% lower administrative burden in 2007 compared to the level of administrative burdens at 31-12-2002. This level of administrative burdens has been measured by applying the Standard Cost Model. The reduction policy and the success of this policy must therefore be regarded within the context of the Standard Cost Model to measure administrative burdens. However, it must be observed that the implementation of the SCM was only one of the reasons for the success of the Dutch policy. Of utmost importance was also the political decision to set a 25% reduction target. Each of the nine ministries involved were called to comply with a set target. In principle, no distinction was made as regards the nature of the regulations: however some ministries could prove they already had reduced the administrative burdens of their regulations. These ministries were allowed to comply with a lower target. This implied a higher target for other ministries, because the overall target was kept at a 25% reduction of total administrative burdens. This fairly simple reduction target was laid down in the coalition agreement that formed the basis of the second Cabinet Balkenende. This made every minister responsible for its reduction policy. Another important feature of the Dutch system is the governance structure that was put in place to achieve the set targets. The appointment of the Minister of Finance as the body in charge of meeting the Cabinet objective, in close collaboration with the Ministry of Economic Affairs, and the creation of a central coordinating unit (IPAL) under the authority of the minister of Finance allowed for constant monitoring of the measurement process via the budget cycle: the annual Budget Memorandum, the Ministerial budgets (September), and the annual reports (May). In addition, the role of Actal, the independent advisory board on administrative burdens established in 2000, is also very important. Actal advises the Dutch government on red tape reduction issues, and acts as a watchdog and facilitator of the reduction policy. Before every new law is sent to the Council of Ministers for Cabinet approval, it needs to pass muster at Actal, which scrutinises the enclosed measurement of the administrative burden stemming from the law. On a more general note, it must be borne in mind that the measurement methodology is conceived to make a reduction policy possible. Accordingly, the underlying assumption that the administrative burden can be measured on the basis of the integrated cost of the hours spent seems suited to the purpose. Moreover, the measurement yields point estimates, which do not account for the observed variation in the time spent to comply with the IO/DRs. These features of the SCM make it difficult to use the measurement outside the context of a reduction policy. The Dutch Court of Audit implicitly criticised this starting point with his remark that businesses do not fully perceive the reduction. The reason for this is that the rules that are being scrutinised are often less onerous in practice than the government presumes, and – 50 – businesses perform certain administrative tasks also for other purposes, or would still have to provide the information concerned to other parties, apart from central government.59 Moreover, the Dutch measurement has suffered from the problem of how to account for one-off costs borne by businesses.60 Currently, these costs are not fully integrated in the model: however, when sending documents to the Council of Ministers departments are obliged to provide a measurement of administrative burdens and also to indicate and measure the one-off costs involved. As a result, one-off costs are somehow taken into account, although they are not considered when setting the reduction target for each department. Once these caveats have been taken into account, the overall conclusion that can be drawn is that – within the limits set by the SCM and compared to the politically defined targets – the Dutch reduction policy is being successful. 59 [Reference] – 60 [TK 29515, 135, p. 5] – 51 – 2.2 Denmark In August 2004 Denmark started a full-scale SCM measurement of the administrative burdens for businesses. The measurement is part of the Danish Growth Strategy and was expected to be completed in October 2005; however, its completion was reached only in March 2006.61 In total, 15 ministries were included in the measurement. The timetable was set to measure initially these ministries which were expected to produce the largest burdens, while ministries imposing smaller burdens on businesses were measured at a later stage. The measurement will be updated every year with the administrative consequences of new laws and executive orders, in order to achieve the target of reducing the overall administrative burden up to 25% in 2010 as required by the Danish Growth Strategy. Businesses should transmit resources from information obligations to innovation and product development. This strategy should enable enterprises to create growth and employment. To make the reduction process transparent, the government decided that every year the amount of administrative burdens which is reduced by each ministry will be reported to the public until the target is reached in 2010. The measurement exercise is also debated in the Danish Parliament, although the Parliament’s interest seems to lie more in the thematic framework of better regulation than in details of the measurement results. Nevertheless, the publication of the results put pressure on the ministries which are competing in reaching the results early. If ministries lie behind their targets or even increase the burden, the public – but also the prime minister or the minister of finance – can also react. 2.2.1 Methodology 220.127.116.11 Definitions In their application of the Standard Cost Model Denmark used the following definitions: • Administrative costs are the costs regarding the administrative activities that businesses have to carry out in order to comply with the information obligations that are imposed through official regulations.62 • Administrative burdens are that part of the administrative costs that the businesses sustain simply because it is a requirement from officialdom.63 • An information obligation (IO) is the passage in a law or regulation text that requires a business to provide or draw up information, and make this available, i.e. 61 „The Danish Growth Strategy“, August 2002, published by the Danish government 62 See Manual for conducting Standard Cost Model measurements, 8 63 See Manual, 9 – 52 – a duty that the business cannot avoid without entering into a position of opposition to the law.64 • Data requirements65: Each information obligation consists of a range of different information that the businesses shall provide in order to be able to comply with the IO. These are called “data requirements”.66 • Cost parameters are the expense variables that are associated with the individual administrative activities. The included cost parameters are differentiated in internal cost parameters (time, hourly rate, overhead), external cost parameters (time, hourly rate) and acquisitions.67 • Private businesses are understood to mean units (physical persons, companies and other legal entities), that produce and/or supply goods/services under market conditions with the objective of generating profit for the owners.68 • Normally efficient business: This is understood to mean businesses within the target group that handle their administrative tasks in a normal manner, neither better nor worse than may be reasonably expected.69 18.104.22.168 ABC-categorisation According to the ABC-categorisation: Category A Data requirements that are exclusively and completely a consequence of EU rules and other international obligations. The international rules describe which information businesses have to produce. Category B Data requirements that are a consequence of EU rules and other international obligations. The purpose will be formulated in the international rules, while implementation (including formulation of the specific data requirements) will be left to the member states. The international rules do not describe which information businesses have to produce. Category C Data requirements that are exclusively a consequence of rules formulated at national level. The Standard Cost Model was used to show the origin of the administrative burdens. The hired consultancy firms70 did the breakdown of legislation into information obligations and data requirements and submitted it to the responsible ministry. The 64 See Manual, 10 65The Danish Manual uses the term "messages" instead of "data requirements".This originally used denomination was replaced by "DR". For better understanding and uniformity this report only uses the term "data requirements". 66 See Manual, 10 67 See Manual, 12 68 See Manual, 15 69 See Manual, 12 70Denmark hired three consultancy firms to do the baseline measurement. Each consultancy firm had a specific contract for specific areas / laws. At the moment, there is only one consultancy firm (Ramboll) to work on further measurements. This firm is / was also involved in the measurements of other countries. – 53 – ministry validated the list of IOs to ensure its completeness and then proceeded to classify the data requirements to the ABC-classification, to clarify whether the burdens stem from Danish regulation, EU or any other international legislation before identifying any reduction targets. To distinguish between EU-legislation and international rules the categories A and B have been further subdivided into: • Category A/B-EU-Directive • Category A/B-EU-Regulation • Category A/B-International The classification is not only useful to draw a picture on the actual origin of administrative burdens, but also allows the identification of ministries that can exert greater impact on the reduction of administrative burdens. Ministries with high administrative burdens in the categories A and B are not able to reduce the burden directly, but in category C ministries could simplify national rules and so it is easier to fulfil reduction targets. The Danish manual also mentions a longer-term purpose of this classification, i.e. to identify the most burdensome EU directives as a basis for proposing changes at EU level. As will be explained in more detail below, the Danish measurement made it possible to calculate the burden of specific EU-directives. 22.214.171.124 Information obligations, data requirements and administrative activities As already mentioned, in Denmark IOs were subdivided into data requirements. The aim is the development of a standard cost catalogue of various types of DRs. The DRs are classified into two types of categories, “data requirements – process” and “data requirements – content”. The first category is subdivided into three categories that describe the process of complying with the DR (e.g. manual process, partially automated process, predominantly automated process). Figure 6 shows a decision tree how to classify DRs based on the process of complying with the DR in the business. – 54 – Figure 6 – Decision tree for classifying data requirements - process One-off or situation - One-off or situation - related DRs – YES related DRs – information constructed Manual process Manual process information constructed each time each time NO information is generated information is generated Predominantly YES Predominantly by means of systematic by means of systematic automated automated processes in the processes in the business process process business NO Some of the information Some of the information YES Partially Partially is generated by means of is generated by means of automated automated systematic processes systematic processes process process Source: Manual for conducting Standard Cost Model measurements, 28 The second category is subdivided into seven categories that describe the content of the DR (e.g. business data, production data etc.). Figure 7 illustrates the classification of DRs according to the content of the DR. – 55 – Figure 7 - Decision tree for classifying data requirements - contents DR relates to fixed DR relates to fixed YES data about the data about the Business data Business data business business NO DR relates to DR relates to Purchasing Purchasing input/output and/or input/output and/or purchase/sale of purchase/sale of goods goods Sales DR relates to DR relates to YES Sales produced goods produced goods DR relates to data DR relates to data about product itself Product data Product data about product itself YES NO DR relates to DR relates to production of goods Production data Production data production of goods DR relates to actual DR relates to actual YES production of goods production of goods in the business in the business DR relates to DR relates to personnel who personnel who Personnel data Personnel data NO produce goods produce goods YES DR relates to DR relates to YES processed data from processed data from Accounting data Accounting data above categories above categories Source: Manual for conducting Standard Cost Model measurements, 30 The administrative activities the businesses have to carry out to comply with the individual data requirement are standardised into 16 activities, in line with the international SCM Network’s Manual (see Table 2 in section 1 above). Importantly, Denmark also measured IOs that would be complied with even if compulsory regulation would be removed. Also IOs generated by optional schemes (e.g. asking for a subsidy) were covered by the measurement if such obligations are considered as conditions for being on the market. So, like other countries, Denmark measured costs and not burdens. The identification of IOs as obligatory or voluntary obligations was carried out by consultants in collaboration with the involved ministry. IOs were differentiated into three categories71: • IOs relating to subsidies • IOs relating to certificates • Other IOs relating to regulation requirements The consultants had to integrate the IOs into this scheme. Creating further categories or proposing new categories was not impossible but required the approval of the Danish Division for Better Business Regulation. 71 See Manual, 38 – 56 – Denmark was also confronted with the problem of demarcation, especially in cases where the same administrative activity can be attributed to two or more rules, as different rules require businesses to provide the same information. Without an exact demarcation, double counting problems may arise. In Denmark, when two regulations contain the same IO, the cost of producing the information is divided equally between the two rules. The Danish manual highlights the importance of “a clear demarcation and definition of the area of regulation where administrative costs are to be measured.”72 Furthermore, consultants were asked to integrate information on related regulations in the data set for the business-related laws. They also have to describe the connections between different laws and their overlapping in obligations. 126.96.36.199 Identification of relevant business segments Segmentation of businesses was carried out on a case-by-case basis. The most important segmentation criteria were firm size (number of employees), industry sector, whether the businesses report digitally or manually and whether businesses outsource the task to comply with the information obligations or not. The segmentation was “target-oriented”, all segmentations exhibit a trade-off between cost effectiveness (“ask as less firms as possible”) and the need for accuracy (“ask enough firms to get accurate results”). If segmentation does not fit reality, the firms in each segment ultimately differ from each other, and too many firms would have to be asked to get representative results. If the segments are well chosen the firms tend to be equal and so a small sample of firms will represent the whole segment. The Danish segmentation therefore tries to approach segmentation on a case by case basis to hold the costs low (each further business segment requires at least five additional interviews with businesses), and to reach accurate results through proper segmentation. If the focus of the measurement does not lie on sectoral impacts, it might not be as worthwhile to analyse sectoral impacts of each information obligation. In line with the prototype SCM, a population was then selected for each segment. As already mentioned, the IOs are subdivided into data requirements. DRs have a rate instead of a population. The rate indicates the proportion of businesses complying with the IOs that have to comply with the given DR. In some cases only some businesses have to comply with a specific data requirement while all businesses also have to comply with the rest of the DRs. The frequency indicates how many times a year an information obligation has to be complied with. As regards the data sources used to identify the relevant population, rate and frequency, Ministries use already available information on businesses registration, statistical materials, and surveys: in some cases, where it is not possible to obtain the necessary information, a qualified estimate was made. The Danish manual highlights the importance of documenting the source of the information to be able to reproduce the information in the same way in subsequent measurements. 72 See Manual, 38 – 57 – 188.8.131.52 Interview guide No standard interview guide was used in the Danish measurement. Accordingly, the consultants had to prepare an interview guide for individual cases, which had to be approved by the Commerce and Companies Agency. In addition to gathering quantitative data, the interviews are used to shed light on qualitative aspects, for example73: • Proposals for rule simplification • Irritation burdens • Best practice in businesses • Whether businesses use the information in other contexts • Whether businesses just prepare information for the government or whether they also use it themselves As a result, Denmark includes the opinion of businesses in their measurement, especially as regards the way in which businesses perceive given pieces of regulation, or whether a specific IO is regarded as especially irritating/onerous. Furthermore, the question whether an administrative activity would be carried out even absent regulation helps to differentiate between administrative burden and administrative costs. Although it was planned to ask the businesses whether they would continue producing the information when the information obligation is repealed, it was not realistic to do so. The businesses were not able to answer the question whether they would continue producing the data requirements or not because most of them did not think about it before. Only in some cases where the burden percentage showed to be very low it was qualitatively reported. If such a differentiation had been reported for each piece of legislation, then the Danish model would also be able to report net costs (as in the UK model, see below, Section 2.3). 184.108.40.206 Cost parameters/Overhead tariff In line with the prototype SCM, the cost parameters were subdivided into internal costs, external costs and acquisitions. However, acquisition costs are measured directly as part of the overhead if they are not solely used for the specific information obligation. Otherwise, the costs are calculated per year. Where an acquisition has a service life of several years a fixed annual cost equivalent to the total cost divided by expected service life is specified for the acquisition. Usually it is not necessary to obtain information on acquisition costs from businesses because “normal” prices of these acquisitions could be obtained on the market. To calculate internal costs the pay index for the private sector (LON02) from Statistics Denmark was used to provide the hourly pay of personnel groups. Hourly pay 73 See Manual, 49 – 58 – represents average rates irrespective of gender. The personal groups and their hourly pay had to be used by the consultants to calculate the costs of the administrative activities. Denmark chose one overhead tariff of 25% for all sectors. This calculation was made in connection with their pilot measurement of the Presentation of Accounts Act. The manual provides a list of items included in the overhead of 25%. In case consultants thought that the overhead percentage would differ significantly from the 25% in individual cases, they had to estimate the overhead percentage themselves, motivating it and submitting the percentage for approval by the Commerce and Companies Agency. Acquisitions are included in the overhead, as they may have a long lifetime or may be used to comply with other IOs. According to the experience of the Division for better business Regulation, the overhead would remain the same were the Danish measurement to be repeated in the near future. External personnel groups and the corresponding hourly tariff were also specified before the measurement. As occurred with internal costs, the consultants could choose different tariffs for individual cases, by motivating this choice and securing approval by the Commerce and Companies Agency. 220.127.116.11 Selection of typical businesses for interviews / normally efficient business Whenever a large proportion of the target group is expected to comply with an IO, a random sample of businesses is identified for interviews. A business is normally asked a few initial questions aimed at assessing whether it can be considered as representative (i.e., “typical”) of the target group. It is also possible to use this initial contact as a survey about outsourcing, IT use or other matters that seem to be significant with regard to business’s costs in complying with the IO. To the contrary, if it is difficult to find a sufficient number of affected businesses through random samples, the responsible ministry – which often has very concrete information on the individual businesses covered by the given IO – is asked to choose a sample. Between three and five interviews with typical businesses in each business segment were taken to provide a qualitative insight into the resource consumption of a normally efficient business with regard to the IO in question. The strategy to identify normally efficient businesses drawn up by consultants has to be approved by the Commerce and Companies Agency and the procedure has to be described in the status report to be produced at the conclusion of the data collection phase.74 The status report was used to ensure that the timeframe for the baseline measurement was actually respected. Since the schedule was very tight, it had to be ensured that the procedure chosen by the consultants was reasonable and possible to carry out within the given timeframe. After gathering the data through the interviews, the time consumption for a normally efficient business with regard to an IO was calculated. This process was carried out in 74 The status report is only available in Danish. – 59 – line with the filtering process suggested by the International SCM Network, as illustrated above in Section 1, Figure 4. If this method to select the normally efficient business resulted in too widespread results, further segmentation was done. For example if there were large differences in the time estimates firms have to invest to comply with an information obligation (or with an administrative activity) sectoral segmentation was also considered. If different sectoral implications are obviously mentioned in the rule sectoral segmentation is generally used. Also, if there are digital solutions mentioned in the rule they are included in the segmentation process. Overall, digital solutions as a method to reduce administrative burdens did not bring these great reduction effects as expected. The activities to work on the required IO take in many cases much more time than the mere act of compliance. After calculating the time needed by a normally efficient business to carry out the various administrative activities, the internal costs and the costs of outsourcing are calculated. After approval by the Commerce and Companies Agency, the validated data were scaled up to national level for each individual segment in the analysis. The scaling up follows the SCM formula: time x pay costs (inc. overhead) x population. 18.104.22.168 Database The Danish measurement is recorded in a complex database. The experiences of three pilot projects were used to establish a structure, were not only the data are visible, but also calculations could be done. The structure of the database is illustrated in Figure 8. Furthermore, the database is available for ministries on the Internet. The ministries can only see the data, whereas the staff of the Division for better business Regulation can also work with the data. The smallest data unit reported is administrative activities. On the basis of the administrative activities different aggregates could be calculated, e.g. the administrative burden according to EU-directives, to ministries, to rules etc.. An adaptation of the database according to EU requirements, such as sectoral impact of different information obligations, is possible, but the data to extend the database might have to be collected. – 60 – Figure 8 – The Danish Database model The database provides seven predefined reports, where users (e.g. ministries) could choose: • Total costs per ministry and agency • Total costs for one ministry per agency, laws and executive orders • Total costs for one law/executive order • Total costs per IO and standard activity • Change in burdens over a given period of time • Total costs by national and EU legislation • Top 100 the most burdensome laws and executive orders As regards the possibility to establish an EU-wide database, the Division for Better Business Regulation was rather sceptical. The argument thereof dwells on the actual “focus” of an EU database. Whereas the OECD Red Tape Scoreboard focuses on describing how the burden looks like, the target of the EU should be the reduction of the burden. Therefore, an EU-wide database, where all 25 countries provide the data estimated through an overall baseline measurement would not be useful. As soon as it is possible to highlight the most burdensome areas of EU-legislation through evaluating a few (for example ten representative countries), this might suffice. Hence, the core analysis of the EU-methodology should focus on sectoral implications rather than accurate absolute numbers. This focusing would allow further analysis what – 61 – simplifications in what sectors result in the greatest growth effects or other economic benefits. The structure of the database teaches another lesson: If areas are calculated using a database it is important to clarify what is included in each area. If there are different definitions used a comparison is not possible or useful. Another example is the calculation of the burden caused by EU-directives. The Danish experiences with benchmarking is that it has to be clarified exactly what IOs are requested by a directive. Otherwise different countries may count different data requirements because the implementation of the directives is different. When accurate numbers are estimated a comparison seems to be useful. If differences exist in the results then it might be important to think on purchasing power parities, “per-capita burdens” or similar standardisations to make the numbers comparable. The database is updated every year. Changes in legislation are steadily worked in the database. The estimates therefore assume that the population is frozen, in other words, the burden is not reduced if the number of firms goes down and the legislation remains unchanged. 2.2.2 Content The baseline measurement covers active private Danish businesses where also IOs to third parties (public authorities, employees, consumers, etc.) are measured. To identify businesses, i.a., the business register (the Danish CVR register) was used. The register contains also businesses that are not really active. These and also foreign businesses are not included in the measurement. The measurements covered all business related legislation, i.e., all laws and executive orders that affect businesses. Therefore, the Ministry of Ecclesiastical Affairs, the Ministry for Refugee, Immigration and Integration Affairs, the Ministry of Foreign Affairs and the Prime Minister’s office where not included as their official acts and regulations are not related to business. Another notable exception is laws or executive orders that are considered to involve less than 1000 hours administrative work for all the businesses concerned.75 Consultants had to draw up a list of the business-related rules that are not included in the analysis, including the reasons for their omission.76 A threshold is important if a baseline measurement is done, but if only partial areas are measured a fixed threshold might be irrelevant. All obligatory rules that the businesses have to follow were measured with the SCM measurement. Voluntary rules are partially covered by the measurement. The Danish version of the SCM distinguishes between voluntary rules that the businesses follow because it is necessary in relation to being on the market and, on the other hand, the 75The first considered threshold of 100 hours was modified, but the threshold is not generally binding. It depends on the number of affected firms. If there are only a few businesses affected by a rule the threshold could have been ignored too. 76 The lists are only available in Danish, normally as appendix to each status report. – 62 – rules that the businesses follow without being regarded as necessary. The decision to include a voluntary rule in the measurement was done case by case. In general, most of the voluntary rules were included. 2.2.3 Procedure The Division for Better Business Regulation (DBR) as a part of the Danish Division foe Better Business Regulation (Ministry of Economic and Business Affairs) is responsible for coordinating the measurement process using the SCM. The DBR must provide an overview of the volume of administrative burdens for every involved ministry, and is in charge of reporting the changes in the level of administrative burdens since the government took office in November 2001. In addition, the Division for Better Business Regulation is in charge of updating the baseline measurement (the database is updated every year) and to perform an impact assessment of new legislative proposals with assistance of the responsible ministries (ex-ante measurement – exactly the same methodology is used in the ex-ante measurement and in the ex-post measurement with the exception that in the ex-ante measurement one-off-costs are taken into account.). There exists no overall budget calculation, but the annual ex-post update of all ministries costs between € 200.000 -300.000. On the other hand, an ex ante measurement is typically completed in one month at a cost of €10,000-20,000 per act. The equivalent of 6-7 staff members are working full- time in the coordinating government unit, and are mostly in charge of ex ante measurements, completing ex-post measurements and overseeing updates of ex post measurements. Moreover, in the nine ministries that cause the greater administrative burdens for businesses, the equivalent of one staff member has been working half-time for approximately 8 months coordinating the ministry’s measurement activity. There exists also an estimate of the costs of outsourcing the measurement (e.g. hiring consultants), namely €2 million. These costs include the costs of outsourcing the measurement in the pilot projects, the baseline measurement and the first update of the database. The Danish SCM manual recommends that it is important, especially in the preparatory analysis, that • the Commerce and Companies Agency is involved on an ongoing basis as regards methodological and technical choices; • the responsible ministry is involved on an ongoing basis in ensuring technical quality; • relevant experts are involved in evaluating and validating the chosen classifications and specifications.77 These recommendations should validate the information produced by the analysis and guarantee uniformity. 77 See Manual, 23 – 63 – Denmark collected the data using – in order of importance – face-to-face interviews, expert assessments, telephone interviews, focus groups, business panels and consultancy studies. The decision on how to collect the data depended on the complexity of the analysed area. Businesses were asked in telephone interviews. When an area seemed to be too complex to be captured through telephone interviews – which was the case in most of the areas – , other forms were chosen. In general, around 80 percent of the administrative burdens are measured through face-to-face interviews. In the interviews, businesses are asked which of the 16 standard administrative activities they have to carry out to comply with a data requirement, and how much time is spent on each activity. Accordingly, it is possible to evaluate the individual data requirements that make up an IO. The involved ministries and business organisations are strongly involved in the measurement process. The data collection (business interviews) is carried out by consultants. The whole measurement was carried out as a cooperative endeavour between the DBR, the relevant ministries, one consultancy firm and also relevant business organisations, depending on the sector affected by the measurement. Expert assessments were used instead of business interviews in cases when the number of affected businesses by an IO is very small. In these cases, specialists with considerable knowledge of the field are involved in the quantification of the administrative costs for a normally efficient business. Denmark hired their consultants through a public tender. The consultants were hired for the whole timeframe in which the baseline measurement was completed. After the baseline measurement, a new public tender for the annual updates was run. The quality of the work of the consultants was ensured through validation by line ministries and the Division for Better Business Regulation. Table 14 below shows how the work was split between the involved units. – 64 – Table 14 – Who does what in the 15 steps of the SCM, Denmark Consultant(s) Coordinating Ministry(ies) Central Phase Step Unit Description Identification of business-related regulation to be included in Phase 0 0 X X X the analysis Identification of IOs/DRs/administrative activities and 1 X X classification by origin 2 Identification and demarcation of related regulation X X 3 Classification of information obligations by type X X 4 Identification of relevant business segments X X Phase 1 5 Identification of population, rate and frequency X X 6 Business interviews v. expert assessment X X 7 Identification of relevant cost parameters X X 8 Preparation of interview guide X X X 9 Expert review of steps 1-8 X 10 Selection of normally efficient business X 11 Business interviews X Phase 2 12 Completion and standardisation for each segment by activity X 13 Expert review of steps 10-12 X 14 Extrapolation of validated data to national level X Phase 3 15 Reporting and transfer to database X X Experiences with consultants in the Danish measurement showed that it might be useful to hire consultants for long periods, as the initial investment to get familiar with the SCM tool is substantial. Therefore, a larger timeframe to complete the measurement seems to be less expensive, as learning effect determine the emergence of economies of scale overtime, thus improving the efficiency and effectiveness of activities carried out by consultants. Furthermore, the Danish experience shows that a consortium of or more than one consultancy firm might cause problems, e.g. rivalry between the consultants. Denmark hired three consultancy firms for doing the baseline measurement and one (the now hired consultancy firm was also involved in the baseline measurement) to do the annual updates. It seems to be efficient if consultancy firms are involved not only in one country. So, the countries could benefit from experiences made by the consultants and, furthermore, the consultants may be more efficient and do not need any preparatory education. 2.2.4 Targets Denmark set an overall reduction target of 25% for all ministries before starting the measurement. The target was set politically and is the backbone of the exercise that ministries are focussing on. There are no sector specific targets set. The timeframe is – 65 – from November 2001 (the zero point measurement) to 2010. As already recalled, the Danish government launched a general Growth Strategy for 2010; the reduction target was therefore set within this strategy. A more refined calculation of the reduction potential for specific pieces of legislation was anyway undertaken in Denmark. Table 15 below reports the estimated reduction potential for three major pilot measurements (Financial Statements Act, Law of Statistics in Denmark, VAT Law). Table 15 - Simplification potential in the three measured areas: *It is important to emphasize that the reduction potential has not yet been realised. However especially the administrative costs from the VAT Law have been reduced significantly. After evaluating the baseline measurement, the seven most burdensome Ministries in Denmark are responsible for around 95 per cent of the administrative burdens for businesses in Denmark. These ministries are: • Ministry of Economic & Business Affairs • Ministry of Taxation • Ministry of Family & Consumer Affairs • Ministry of Employment • Ministry of Environment • Ministry of Food, Agriculture &Fisheries • Ministry of Justice Furthermore, around 90% of the burdens are concentrated in the 10 most burdensome regulations within each Ministry. In order to realize the 25 percent reduction target the focus is therefore on the seven most burdensome Ministries and the 10 most burdensome regulations within each Ministry. These seven Ministries have delivered in September 2006: • A long term simplification plan showing how the Ministry will reduce the administrative burdens by 25% in the coming 4 years (until 2010). The Ministries have set a reduction target for each of the top 10 most burdensome regulations. • A short term simplification project plan with a detailed description of minimum two concrete simplification projects (from the top 10 list) that the Ministry will initiate in the coming 6 months. – 66 – The simplification action plans are going to be approved by the Governments Economic Committee in September and the Governments Coordination Committee in October 2006.78 2.2.5 Results The breakdown of administrative burden to report them was done by ministries, agencies, primary and secondary laws. If the same regulatory requirement can be attributed to more than one ministry the costs are divided equally between the ministries. An overview of administrative burdens of each evaluated ministry is shown in Table 19 below. Overall administrative burden = € 4,172,400,000 (latest update 2005) (2.13% of GDP) Per capita burden = € 771.04 (5,411,405 citizens per 2005) Table 16 below shows central aspects of the Danish measurement. As shown in the table, the largest share of the costs has its origin in national legislation (57%). There were 263 laws and 1,100 executive orders included in the measurement, which were split up into 5,279 information obligations. Table 16 – Central figures from the zero point and baseline measurements Administrative costs in 2001 (€ million) 4,392.9 Administrative costs in 2004/05 (€ million) 4,326.3 Development from 2001 to 2004/05 in per cent - 1.5 % Origin of the administrative costs in per cent: A (originates directly from international regulation) 28 % B regulation (international origin, but implemented nationally) 15 % C regulation (national origin) 57 % Laws included in the measurement 263 Executive orders included in the measurement 1,100 Information obligations 5,279 Data Requirements 14,546 Number of face-to-face business interviews 1,100 Hours of face-to-face business interviews 2,902 Source: www.administrative-burdens.com/ 78 See www.administrative-burdens.com – 67 – The origin of the burden can be divided A, B and C. Therefore a table of the top-20 areas exists. In most cases they can trace the directives that are implemented by the national legislation. An estimate of the share of the administrative burden originated by non-EU legislation in category B in Denmark is less than 1%, in category A less than 5%. Table 17 – Administrative burdens by areas and ABC-classification – 68 – The estimates on the burden caused by the EU are available by ministries (compare Table 18 below). The most burdensome ministries according to the EU-caused burden are especially Economic and Business Affairs and Family and Consumer Affairs. Table 18 – Top-5 most ‘EU-burdensome’ ministries/areas in Denmark Furthermore the Danish measurement allows to calculate the burden on Danish businesses caused by individual EU-directives. An exemplary list of burdensome directives is shown in Table 20 below. The procedure to get the most burdensome EU- directives started with calculating the most burdensome areas (presented in Table 17). In these areas further analysis were done. The identification of the most burdensome directives was easy if the implementation into Danish law was done through one single national rule. A more complicated case occurred where one Danish law implements more than one EU directive. Therefore the origin of the information obligations of the Danish law had to be identified. The most complicated case was, – 69 – when more than one EU regulation is implemented through more than one Danish rule. These calculations were not possible in the short timeframe of this pilot project. In general, it might be possible to do it, but it takes time, because different experts in the ministries would have to be asked to work on what information obligation requested in which EU-directive can be found in what Danish rule. Therefore, to work on an EU- methodology, the Division for better business Regulation highlights the need of exact definitions or demarcations, what information obligations have to be measured by the member states to guarantee that all countries measure the same. Otherwise, the results are not comparable. – 70 – Table 19 – Results of the Danish measurement by ministry – 71 – Table 20 – Most burdensome EU Directives in Denmark – 72 – According to the Danish measurement and furthermore on the calculations of most burdensome EU-directives, the Division for Better Business Regulation (DBR) is working on proposals, which can be followed through quick action to reduce administrative burdens caused by the EU (“low-hanging-fruits”). These upcoming proposals reportedly predict significant reduction potentials, but must be considered as a work in progress, not as official statements. 22.214.171.124 The financial effect of administrative reductions The SCM method is used in Denmark as basis for calculating the effects of administrative reductions on growth. Therefore, it is examined what industries are affected and what effects the reduction of administrative burdens and, furthermore, the release of resources has on growth. The resources used for complying with information obligations, for example employees who worked on completing forms, can be used for productive work in the same or in other companies. The growth effect that administrative reductions have on value growth is expressed in a multiplier indicating the relation between the size of administrative reductions and additional growth. To investigate the multiplier effect the “Copenhagen Economics Trade Model” (CETM), which is shortly described in Figure 9, is used. Figure 9 – Description of the Copenhagen Economics Trade Model (CETM) The model is a static used general equilibrium model which describes Denmark as a small open economy. The data incorporated in the model is drawn from the GTAP6 database (see www.gtap.org) with 2001 as the basis year. In principle, it is a global model, which includes a detailed description of the EU25 countries. However, in the calculation of the Danish multiplier, the model focuses on the Danish economy and treats the EU15 countries and the rest of the world as two areas. The model includes a detailed distribution between industries, and therefore, it is useful in an analysis such as this one, where it is important to describe how changes in one national industry affect other national industries through diverted trade effects. On the production side, it is assumed that there is profit maximisation and total competition on all markets. All companies have the same basic cost structure. The production in each industry is performed through a combination of capital, manpower and consumption of semi-manufactures. First, the companies choose the optimal amount of capital and manpower, which is then put together with semi-manufactures. It is assumed that the companies substitute between manpower and capital with an elasticity of 1, while there is a steady relation between value growth and semi-manufactures. The consumption side of the model is described through a benefit-maximising representative household, which has a steady supply of capital and manpower. Capital and manpower are fixed at a general level, but can move between industries. Source: Copenhagen Economics Administrative cost reductions are assumed to enable companies to reach the same production level with less manpower and so increase labour productivity. In the long – 73 – term it results in increased production of goods and services and additional value growth. The higher level of production pushes down prices and increases the average wage rate. The calculation of the multiplier effect is based on a realisation of administrative reductions of DKK 1 billion distributed between industries in the economy in accordance with their share of total number of employed people in Denmark. The simulation indicates an increase of labour productivity and a corresponding increase in total value growth of almost DKK 1.4 billion in the long term. Accordingly, the multiplier between the change in value growth and the size of the administrative reductions lies at about 1.4. Although the administrative reductions of DKK 1 billion are only a numerical example the overall multiplier remains unchanged if the amount of reduction changes. The Danish calculations include also sector specific reductions, which correspond to different multipliers, depending on the differences of the industries and also how the different industries affect other parts of the economy. One of the major aspects, how the multiplier looks like, is the export share of the industry affected by administrative reductions. The higher the export share of the affected industry, the higher is the multiplier. The multiplier also depends on who buys the products of the affected industry. If other industries buy them the reduction causes further increases in production and so leads to a higher multiplier. If households are the primary buyers, there are no positive production effects on the rest of the economy. According to these arguments the Danish calculations of administrative reductions in specific industry sectors are illustrated in Table 21 Table 21 - Multipliers for the entire economy and individual industries – 74 – Source: Calculations on the basis of the Copenhagen Economics Trade Model (CETM). Note: The multiplier is calculated on the basis of the change in value growth for the entire economy divided by the size of the administrative reductions. Note: The calculations presuppose that administrative reductions increase labour productivity permanently. It is also presupposed that administrative reductions of DKK 1 billion are distributed on the industries of the economy in correspondence with the distribution of employment between the industries. 2.2.6 Summary The Danish implementation of the SCM led to a successful measurement, in which uniformity of results was guaranteed through reiterated control and central coordination. The measurement process, especially the preparatory stage, was formed through a top-down approach, where the central unit prepares standardised lists of cost parameters, administrative activities etc. Every deviation from the proposed standards had to be approved by the Commerce and Companies Agency. According to its experiences the Division for better business Regulation expects the SCM method as it is used in Denmark to be applicable on EU level. The methodology does not differ between the countries compared in this project. Only some decisions were differently made (e.g. whether the burden on citizens is measured or not). Furthermore, some changes have to be made nonetheless, because for example data collection methods strongly depend on national peculiarities. Anyway, in most cases it might suffice to measure only in e.g. ten countries and to identify only the most burdensome areas or directives to come up with reduction targets. The question is how to extrapolate the data from national to EU level and whether it is necessary to investigate all or only some “typical countries”. Furthermore, it would have to be clarified: what are these “typical countries” and how important is the exact figure of the burdens? Maybe it is only important to know the relative burdens to focus on what regulations should be repealed or changed. On the other hand, if the costs of a regulation are known, it is normally possible to assess whether it would be worth working on the reduction of the corresponding burdens. The Danish measurement uses a subdivision of IOs, i.e. DRs, which seems to be a useful move to capture the complexity of administrative burdens accurately. However, the level of detail reached by the Danish government may prove excessive in larger countries, in which the measurement of IOs broken down in specific DRs would probably lead to measurement costs skyrocketing. Most importantly, the Danish measurement allows, in most but not in all cases, for calculating administrative burdens generated by individual EU pieces of legislation. Sectoral analyses, i.e. identifying what sectors are most hit by an individual EU piece of legislation, is however not always possible; in particular, it is not possible to classify the sectors according to international industry classifications (e.g. NACE codes) at the moment, but it would be possible after updating the database with the required data, which might be more or less expensive. To the contrary, producing data segmented by firm size is possible, as the measurement explicitly considered this aspect in cases were – 75 – it was expected to be a relevant criteria for the time spent complying with the information obligation. Moreover, the fact that businesses were asked which obligations are particularly irritating or onerous might help in identifying priority areas for simplification or reform, i.e. the so-called “quick wins” or “low-hanging-fruits”. The main lessons that can be learnt from the Danish case study are the following: • Detail: subdividing IOs into DRs is certainly useful • Standardisation of cost parameters (top-down approach) is very important to secure the uniformity of the measurement, also in case of measuring EU-directives it has to be clearly defined what information obligations have to be measured by member states. • The Danish model allows for calculating burdens according to individual pieces of EU legislation, at least in the majority of cases • Sectoral analysis: sectoral differentiations are not included in the Danish measurement: in general, only differentiation in firm size is used. In some cases segmentation by business sector was also considered, but such segmentation was not carried out according to international classifications (NACE). • Qualitative information was collected during the measurement: businesses were asked whether specific IOs were particularly irritating or onerous, and this seems to allow for the collection of useful suggestions for reform (i.e. “low-hanging- fruits”). For this purpose, no overall standardised interview guide was drafted in advance. • Accordingly, the Danish database allows for an efficient identification of simplification priorities (“low-hanging-fruits”) • The segmentation of categories A and B into three sub-categories (EU-Directive, EU-Regulation and International) makes the Danish model particularly suited for communication with the EU SCM, although some adjustments will be needed. • The Danish database allows extensive analysis of e.g. low-hanging-fruits – the database seems to be one of the most important aspects of a successful and useful (e.g. to do further calculations and future use) measurement. It is possible to adapt the Danish database to European level and also to include sectoral segmentation, although the data have to be collected. • The Danish Division for better business Regulation advises to focus on similarities and not differences of the SCM approaches to establish a common EU methodology. • To provide a good measurement a project on measuring administrative burdens needs strong political backing. The Danish project benefited from the interest of the prime minister and also of the minister of finance. – 76 – • The Danish experiences on hiring consultants suggests to concentrate on one consultancy firm and to hire it for a long period. Also the timeframe of the measurement should be large enough to utilise these benefits. • The measurement of administrative burden should focus on reduction targets and on identifying reduction potentials rather than on establishing an exhaustive database with absolute numbers of administrative burdens, maybe according to industry sectors. What areas/directives/rules are most burdensome is more important than how large these burdens are in absolute terms. • A distinction between administrative burdens and administrative costs was tried through collecting qualitative data. This approach seemed to be impossible or to complex since most of the asked firms had not thought about it. • Digital solutions do not have a large reduction potentials as was expected. Without simplifying activities the firms have to do to comply with the obligations digital solutions do not result in high reductions, because the transmission of required data is often not as expensive as the process of calculating them. • The more complex the measurement process, the higher is the importance of the preparatory stage: especially in the case of the EU there exists the need of standardisations and definitions what has to be measured by the member states. – 77 – 2.3 The United Kingdom Amongst the EU25, the UK has probably the most consolidated tradition in the field of better regulation, especially as regards the use of compliance cost assessment (since 1985) and regulatory impact assessment (from 1998).79 The adoption of the SCM was recommended by the Better Regulation Task Force in the “Less is More” publication in March 2005, and was then implemented in the ambitious Better Regulation Action Plan launched in June 2005 by the Chancellor of the Exchequer, Gordon Brown. The Action Plan was aimed at embedding a risk-based approach in regulatory decision-making, announced as providing for “no inspection without justification, no form filling without justification, and no information requirements without justification.”80 Under the prescriptions of the Action Plan, Departments started measuring the total administrative burden imposed upon businesses (including the costs of form-filling, undergoing inspections and complying with data requirements).81 Moreover, businesses were invited to propose areas of regulation and administrative costs that are needlessly burdensome, also through a new online portal named www.betterregulation.gov.uk.82 The Panel for Regulatory Accountability within the Cabinet Office is in charge of setting targets for burden reduction for each department by the 2006 pre-budget report. Already in November 2005, some departments started preparing ‘simplification plans’ by reviewing existing legislation and taking into due account proposals coming from businesses as well as guidelines contained in the Hampton review and in the BRTF’s “Less is more” document.83 The Legislative and Regulatory Reform Bill, introduced to Parliament on 11 January 2006, completed its Committee stage in the House of Commons on 9 March 2006. The bill mandates new fast tracks for repealing or amending outdated and unnecessary legislation, sets out the powers to merge regulatory bodies and introduces new tools aimed at promoting the use of a risk-based approach as recommended during the Hampton review. Entry into force of the Bill is scheduled for summer 2006. Detailed deregulatory proposals – mostly based on suggestions by businesses – will then be 79The CCA procedure was launched with the White Paper, Lifting the Burden, Cmnd 9571 (1985). For an overview, see A. Renda, Impact Assessment in the EU. The State of the Art and the Art of the State, CEPS Paperback Books, Brussels, 2006. 80The Action Plan followed Hampton Review on administrative burdens, which had suggested the adoption of a risk- based approach as well as massive consolidation of UK regulators. See e.g. “Brown pledged to Cut Business Red Tape”, The Guardian, 24 May 2005. As a result, Chancellor Brown announced the plan to reduce UK regulators from 31 to as few as 7. See Cabinet Office, A Bill for Better Regulation: A Consultation Document, 25 July 2005, Chapter 2. See the press release of the Cabinet Office, “Cabinet Office kicks-off project to measure and cut red tape on business”, 81 CAB 043/05, 15 September 2005. By 13 July 2006, the portal had received over 100 proposals, with an additional 150 being received through other 82 methods. 83 See the Treasury’s Press Notice of 24 May 2005, “Chancellor Launches Better Regulation Action Plan”; and BRTF, Less is More. – 78 – included in a further ‘Deregulation’ bill to be submitted to Parliament by the second session of 2006.84 In launching the consultation phase on the Legislative and Regulatory Bill, on 20 July 2005, the Government announced that it had accepted all eight recommendations issued by the BRTF in the “Less is More” report, and decided to focus on the following three main objectives: • regulating only where necessary and in a light touch way that is proportionate to risk; • setting exacting targets for reducing the cost of administering regulations; and • rationalising the inspection and enforcement arrangements for both business and the public sector. In January 2006, the BRTF was replaced with a Better Regulation Commission, whose main role is to advise and challenge government departments on regulatory reform issues and to scrutinise departmental plans for regulatory simplification. By the end of 2006, the National Audit Office will start reporting to the Parliament on departments’ performance in implementing the new risk-based methodology and in effectively reducing the burden of regulation, as well as on regulators’ performance against the recommendations of the Hampton Review.85 As was also recalled by the Rt Hon John Hutton MP, then Chancellor of the Duchy of Lancaster and Minister for the Cabinet Office, in launching the consultation on the new Better Regulation Bill, “[t]he potential economic gains from stripping away unnecessary regulation are enormous”. Hutton also recalled estimates by the BRTF that the new approach “could boost British national income in the long term by 1% of GDP – a huge gain of around £10 billion for the UK economy”.86 The BRTF also calculated that adopting the SCM for measuring administrative burdens and then targeting a 25% cut in such burdens over four years could reduce direct regulatory costs on businesses by £7.5 billion, yielding a £16 billion increase in the UK GDP in the medium term.87 As of now, however, targets were set only for three departments (DTI, DEFRA and HSE): in all cases, the target chosen was a 25% reduction by 2010. The baseline measurement started in September 2005 and the results are expected in the second half of 2006. Sixteen departments were involved in this measurement. In the 84 See Cabinet Office, A Bill for Better Regulation. 85See the Cabinet Office’s press notice, “Cabinet Office kicks-off project” and the Cabinet Office’s Consultation Document, A Bill for Better Regulation. 86See the Cabinet Office’s press notice, “Transforming the Regulatory Landscape – Launch of a Consultation on a Bill for Better Regulation” (available at http://www.cabinetoffice.gov.uk/regulation/news/2005/050720_bill.asp – visited 2 August 2005). The approach taken by the UK Government has been termed ‘one in, one out’, to express the need to “encourage a better balance between the introduction of new regulation and deregulation”. See speech by Sir David Arculus, Chairman of the Better Regulation Task Force, to staff of the Financial Services Authority, 29 June 2005 (available at http://www.fsa.gov.uk/Pages/Library/Communication/Speeches/2005/0705_sda.shtml – visited 3 August 2005). 87 See the Statistical Appendix to the Less is More White Paper (available at http://www.brc.gov.uk/news/2005/lessismore.asp#stats – visited 29 August 2006). – 79 – meanwhile a separate exercise was done by the tax authorities of the UK, the results of which were published in March and April 2006.88 As regards the UK implementation of EU legislation, in the first half of 2006 Lord Neil Davidson, the Advocate General for Scotland, led a comprehensive review (the “Davidson review”), aimed at ensuring that domestic implementation does not result in unnecessary regulatory burdens as a result of gold-plating or double-banking. The Review launched a call for evidence between 3 March and 25 May 2006, and a summary of responses was published on 19 July 2006.89 Following further analysis, the Review will report with conclusions and recommendations by the end of 2006. The final report will set out specific proposals for reducing any unnecessary burdens (administrative and policy costs) in the stock of EU-derived legislation, as well as wider lessons for best practice implementation of European legislation in the future. Below, we describe in detail the peculiar features of the UK variant of the SCM, and the evidence gathered so far in the measurement exercise. 2.3.1 Methodology 126.96.36.199 Regulations covered by the measurement The UK baseline measurement covers central government regulations. However, such a definition is to be interpreted broadly, and covers, for example, also non-legislative regulation with government backing, such as approved codes of practice. The UK Standard Cost Model Manual defines regulations as rules “with which failure to comply would result in a business coming into conflict with the law or being ineligible for continued funding, grants and other applied for schemes”.90 This definition includes all measures with legal force imposed by central government and other schemes operated by central government, including: • Directly applicable EU Regulations; • Directly applicable sections of Acts of Parliament; • Statutory Instruments; • Rules, orders, schemes, regulations etc. made under statutory powers by Ministers or agencies; • Licences and permits issued under central Government authority; • Codes of Practice with statutory force; • Guidance with statutory force; 88See KPMG, Administrative Burdens- HMRC Measurement Project, Final Report. And Real Assurance, Estimation of FSA Administrative Burdens, June 2006. . 89 €See http://www.cabinetoffice.gov.uk/REGULATION/reviewing_regulation/davidson_review/summary.asp. 90 See KPMG, Administrative Burdens- HMRC Measurement Project, p.52 – 80 – • Codes of Practice/Guidance/Self-Regulation/Industry Agreements with government backing, e.g. “Approved Codes of Practice” • Bye-laws made by central Government. In turn, the definition excludes: • Codes of Practice/Guidance without government backing; • Voluntary agreements between businesses including self-regulation/industry agreements with no government backing; • Obligations arising from local government measures where these are not concerned with implementing central government policy; • Obligations arising from the Common Law (e.g. of an employer to his employees); • Local authority bye-laws, pure local authority regulation; • Devolved areas where policy responsibility has been transferred to the Scottish Executive, Welsh Assembly Government or Northern Ireland Executive. As regards EU legislation, the UK Manual specified that when EU regulations do not require any domestic implementation, and are therefore directly enforceable as such, the regulations only have to be listed and flagged as such, but it is not compulsory to include them in the measurement – although the Manual specifies that this would be in any case “desirable”.91 This is of course understandable, as the UK baseline measurement aims at identifying areas for reduction of administrative burdens that fall under the sphere of influence of the UK government departments. As will be specified in more detail below, this also means that administrative requirements listed and flagged within the category A-EU-Regulation are not necessarily measured. Observed results in any case testify that such burdens have been measured in practice. 188.8.131.52 Scope of measurement The scope of the UK measurement is slightly broader than the scope of the prototype SCM in some respects, and narrower than the scope of the proposed EU common methodology. This is due to the following peculiarities: • The UK baseline measurement covers administrative burdens imposed by legislation on private businesses and semi-private businesses (e.g. charities), but not public authorities and citizens. The UK SCM Manual specifies that “[t]he term business should be taken to mean every sector of the economy, with the exception of the public sections of public administration and public services as defined by the Office for National Statistics for National Accounts purposes. This definition therefore includes business, charities and the voluntary sector”.92 This also means that, unlike what occurred in the Netherlands, public owned businesses which cover their own costs have not been included in the measurement. 91 “There may be some interest in understanding the burden imposed when considering EU simplification plans and it is for such purposes that it may be desirable to estimate the burden of such regulations.” 92 SCM Manual, §3.1.9. – 81 – • In addition, the measurement covers both information obligation to public authorities and to all third parties, including employees, consumers and shareholders, as required by the individual regulation.93 However, when a Department has a large proportion of obligations to third parties, alternative approaches may be chosen after discussion between the department, the BRE and the selected consultants.94 • In line with the prototype SCM, the UK model includes in the measurement also information obligations that would be complied with even absent a specific legal prescription. This means that the UK model measures administrative costs, not just administrative burdens.95 • The UK model also includes information obligations arising from non compulsory measures (i.e. voluntary obligations). Furthermore, in the UK a 100% take-up of the target groups is assumed, in order to focus on the potential burden where high burdens could be a barrier to full take-up.96 • In the UK, IOs with cost-determined reimbursement are included in the measurement, but the level of reimbursement is recorded in order to include such regulations on a net-basis. In this respect, the UK variant of the SCM differs from the Danish one, where these IOs are not included in the measurement; and it also differs from the Dutch model, in which the amount of reimbursement is cleared in the administrative cost measurement.97 As regards the areas covered by the measurement, the measurement involves sixteen departments. A separate measurement was carried out to estimate the tax burdens on UK businesses, and the results were published in April 2006. No specific areas were excluded from the measurement. More in detail, in September 2005 the departments involved submitted a list of 2,266 regulations that would potentially fall within the scope of the measurement exercise. The BRE and the consultant (a consortium led by PwC) then proceeded to scrutinise this preliminary list of regulation in order to check for consistency and avoid duplications, truncations, omissions and cases of misnamed regulations, while at the same time identifying potential demarcation problems, that were solved at a later stage (at Step 2 of the 15 steps of the prototype SCM, see below). 93 See, contra, SCM Network, Difference in the application of the SCM, July 2005. 94 See Manual. 95However, as will be shown below, results in the UK database are reported both in terms of total costs (administrative costs) and net costs (i.e. administrative burdens). 96See SCM Network, Differences in the application of the SCM. In the Netherlands, where there is known to be a very different level of take-up this should be noted during the listing of regulations and/or measurement process. 97See OECD Review of the Standard Cost Model (2005). The SCM seeks to exclude from the measurement process those costs that are already in the public accounts: the focus is on burdens that are uncompensated by government. However, the OECD recommended that reimbursement be not subtracted from administrative burdens in the measurement. Id. at §112. – 82 – As shown below in Table 22, the final baseline contained 1,323 regulations, plus 112 amending regulations. This list was validated with the responsible departments and was further modified during the implementation of the measurement by the consultant, upon approval by the departments and the BRE. Table 22 – number of regulations by department Source: PwC, Technical Summary 184.108.40.206 Focus: one-off v. recurring costs In line with the prototype SCM, the UK baseline measurement does not cover one-off costs. However, such costs are taken into account in the ex ante measurements for inclusion in a RIA. In this case, for example, the cost of reading guidance to get familiar with a new set of regulatory obligations is taken into account in the assessment of costs related to the regulatory options considered. However, although estimating ‘implementation costs’ has become a mandatory step in the UK RIA model since August 2000, in practice such costs have seldom been calculated.98 Of course, in some borderline cases it may be difficult to distinguish between one-off requirements and information obligations that arise with very low frequency, 98 Is the current consultation for the review of the UK impact assessment model, the inclusion of a mandatory assessment of both “administrative costs” and “policy costs” is envisaged. See the consultation document by the BRE, Delivering Tools for Better Regulation, July 2006. – 83 – especially when such frequency is difficult to estimate. Where departments believe a specific regulation imposes administrative costs that might not be considered as recurring, they should ask the BRE to decide if the regulation should or should not be included, providing the BRE with evidence to support their case. In any event, the regulation at hand is included in the list of regulations and is marked as a rare contingent requirement; moreover, its record also contains details of the BRE decision about measurement.99 220.127.116.11 Borderline cases The UK SCM Manual anticipates that in some instances borderline cases may occur in which it may be difficult to decide whether a regulation falls within the scope of the baseline measurement. Reported examples are the following100: • Complaints: the right of businesses to complain about decisions is not an administrative burden and so is not measured in a standard cost analysis. However, in sectors where complaints are very common, the SCM Manual recommends that a qualitative survey is carried out to find out about the parts of legislation that are most difficult to understand and/or lead to filing complaints; • Household v. business: the border between the two is somehow difficult to draw whenever self-employed individuals apply for schemes or subsidies as employees: in this case, as the individual acts as an employee, the corresponding administrative work is not considered in the measurement; • Inspections: the UK better regulation agenda exhibits a strong focus on reducing the cost of inspections, especially since the 2005 Hampton review. In the baseline measurement, only inspections that can be related to a specific IO are included, whereas the cost of broader inspections is not measured; • Litigation costs are not included, unless it can be proved that cases cannot be avoided by fully complying with an IO; • First time compliance must be included in the measurement, but the population is only composed by the new firms in the segment that have to comply fr the first time each year: the need to read guidance documents and get familiar with the new task is included in this measurement. 18.104.22.168 Information Obligations, data requirements and administrative activities The UK model is based on the internationally acknowledged taxonomy of information obligations (IOs), data requirements and administrative activities. An IO is defined as “a duty to procure or prepare information and subsequently make it available to a 99See Manual. As cases are presented to the BRE, decision criteria for the exclusion of such requirements will be developed. Some of these – complaints, household v. business and inspections – are also reported in the International SCM 100 Manual by the SCM Network, at 22. – 84 – public authority, as well as a duty to facilitate the collection or preparation of information by others, e.g. by permitting and cooperating with an audit, visit or inspection. This includes regular requirements to read guidance and updated rules, for example rules which are updated annually. It is an obligation businesses cannot decline without coming into conflict with the law or being ineligible for continued funding, grants and other applied for schemes.” An indicative list of IOs is provided by the UK SCM Manual in line with the Danish and the SCM Network’s International Manual.101 The UK Manual is unclear as regards the need to measure IOs to third parties. The indicative list reported in Table 1 – which is reported also in the UK SCM Manual – includes some IOs directed to third parties (e.g. labelling): however, the Manual also reports that “it has been decided that such obligations should be identified but not included in the baseline. They should be flagged as relating to third parties in the database.”102 But the Manual also specifies that these IOs should, where possible, be included in the measurement. In practice, IOs to third parties have been measured and will be reported: such IOs are more numerous (and burdensome) in some departments, depending on the areas of policy they are responsible for. In the final reports, obligations to third parties and non-third parties will be reported separately. Overall, IOs to third parties are expected to account for approximately 50% of all burdens. As regards data requirements, in line with the Danish Manual, the UK model prescribes that they are classified in two different types of category103: • Data requirements – process. These include: o Manual process: these are either one-off or situation-related tasks that are performed so rarely that the information has to be constructed each time and is therefore produced “manually”. o Partially automated process: for these tasks, up to half of the information needed has already been generated or can be generated easily. As a result, the task can be accomplished with use of IT tools or systemised administrative processes, with consequent time savings. o Predominantly automated process: tasks for which more than half of the information needed is already available or easily generated. • Data requirements – content. This category contains four main subgroups, i.e. fixed data about the business, data related to produced goods (purchasing, sales, products), data on the production process (production, personnel), and accounting data. Finally, as regards administrative activities, the UK initially followed the internationally acknowledged non-exhaustive list of 16 activities, which could be used by departments in describing the actual activities that need to be performed to comply 101 Id. at 24, and Danish Manual at Box 12. See Table 1 in section 1 for the list of examples of IOs. 102 P. 44 103 See also the Danish Manual, pp. 24 ff. – 85 – with the information obligation.104 Such list was then grouped into 7 sets of activities, which enabled the interviews to be undertaken within the required timescale.105 Interestingly, when an administrative activity is found to entail form-filling, this has to be specified and reported, so that the actual amount of form-filling within the broader category of administrative burdens becomes visible at the end of the measurement per department. In the UK, ambitious targets in the reduction of form-filling have been set, for example, by the HSE. 22.214.171.124 Classification of origin While the UK followed the ABC classification and the 7-step model for classifying the origin of regulations, the UK database and reporting formats allow for further subdividing the categories, such that EU legislation appears visible. This is in line with the model adopted in Denmark, as described above in Section 2.2. In particular, as shown below in Table 23, the UK model subdivides category A and B in three main groups: international legislation, EU Directives and EU Regulations. This also means that the UK database may allow for building a separate category comprising all IOs arising from EU legislation. For the (ten) most burdensome regulations, each department must present a breakdown of the results by information obligation, data requirement, firm size and e-government solution, and highlight the parts of the regulation that drive the observed level of administrative costs. Table 23 – UK reporting format on the origin of regulation (A and B) Source: UK SCM Manual, Annex 5, page A47. During the baseline measurement, departments provided information about the origin of each of their regulations on the original lists. Subsequently, however, the consultant and BRE decided that it would be more appropriate to specify the origin at a more detailed level, by classifying information obligations (IOs) and – where possible – data requirements according to this modified ABC classification. This analysis was carried out by an ad hoc team at the BRE, and was then validated with departments between the end of Phase 2 and the start of Phase 3.106 104 Id. At 25. and IPAL (2003), Annex 2. 105 See Table 7 of PwC’s Technical Summary. 106It should be noted that a recommendation was made to departments by the BRE to focus their validation efforts on those IO/DRs allocated to ‘A’ or ‘B’ origins. – 86 – 2.3.2 Procedural/organisational issues The organisational structure of the UK baseline measurement heavily relies on the role played by the Better Regulation Executive (BRE) within the Cabinet Office, responsible departments and the consultant. In particular, overall responsibility for initiating and coordinating the measurement exercise has rested with the BRE. More in detail, BRE’s project team ensured the coordinating the work across departments, including the development and consistent application of the Standard Cost Model (SCM); whereas the overall management and direction of the project has been conducted through a Project Board chaired by the BRE with representation from a number of departments. All departments have been represented at regular meetings of Senior Departmental Leads (SDLs). Departments have been responsible for the provision of information relating to the regulations in scope, the validation and sign-off of data and the sign-off of their respective reports. Wider stakeholder input has been sought from the Strategic Measurement Monitoring Group. As regards the consultant, the measurement task was awarded to a consortium led by PricewaterhouseCoopers LLP (PwC), which included legal experts from DLA Piper and Landwell.107 The consultants’ involvement lasted eleven months. The procedure followed by the UK in implementing the SCM is in line with the prototype 15-step implementation procedure highlighted in Section 1.2. The BRE chose to involve a single consultant for the main measurement (PwC), whereas the separate measurements carried out by the HMRC and the Financial Services Authority each were carried out with the help of one consultant (KPMG and Real Assurance Ltd). Consultants were selected through a mini-competition from S-Cat lists based procurement operated by the Office of Government Commerce. The main measurement involved, as main actors, the responsible departments, the BRE and the consultant. In particular: • the departments were involved in phase 0 (identification of regulations to be included in the analysis), step 1 (only for origin classification) and step 5 (population, rate and frequency). • The BRE, as central coordinating unit, was involved together with the departments in step 1 (origin classification) and step 5 (population, rate and frequency). • The consultant (a consortium led by PwC) was involved in all phases with the exception of steps 9 and 13 (experts review of Phase 1 and 2). • Departmental and Strategic Monitoring Groups, together with external stakeholders, were involved in steps 9 and 13 (experts review of Phase 1 and 2). 107PwC has also drawn upon the expertise of Brigitte van der Burg, Member of the Dutch Advisory Board on Administrative Burden (ACTAL), as technical adviser. Brigitte’s involvement in the Dutch implementation of the SCM has given her a deep understanding of the subject and an insight into some of the practical challenges involved. – 87 – Table 24 – Who does what in the 15 steps of the SCM - UK Consultant(s) Coordinating Ministry(ies) Central Phase Other Step Unit Description Identification of business-related regulation to be Phase 0 0 X X* X* included in the analysis Identification of information obligations, data 1 requirements and administrative activities and X** X** X classification by origin Identification and demarcation of related 2 X regulation 3 Classification of information obligations by type X Phase 1 4 Identification of relevant business segments X 5 Identification of population, rate and frequency X X X 6 Business interviews v. expert assessment X 7 Identification of relevant cost parameters X 8 Preparation of interview guide X 9 Expert review of steps 1-8 X*** 10 Selection of normally efficient business X 11 Business interviews X Phase 2 Completion and standardisation of time and 12 X resource estimates for each segment by activity 13 Expert review of steps 10-12 X*** 14 Extrapolation of validated data to national level X Phase 3 15 Reporting and transfer to database X * Initialisation of the measurement project and checking of lists from departments ** Classification of origin *** Departmental and Strategic Monitoring Groups with external stakeholders. Source: BRE. – 88 – 126.96.36.199 Segmentation During Phase 1 of the UK measurement, only three departments submitted cases in which segmentation would be needed. Segmentation was to be carried out according to three main dimensions: • firm size was considered on the basis of the four categories used by the Small Business Service (SBS), and thus included108: o Micro firms (0-9 employees) o Small firms (10-49 employees) o Medium firms (50-249 employees) o Large (more than 250 employees) However, as reported by the BRE, firm size eventually was almost never used as a criterion for segmentation, except n the separate tax measurement (HMRC), in which a segmentation by size was found to be useful and reliable. In most other cases, the consultant itself observed that segmentation by size did not offer a reliable picture of the population affected. • Industry sector, although not necessary, may be accounted for according to the SIC 2003 classification by the Office of National Statistics. Interestingly enough, during the measurement exercise this type of segmentation ultimately proved more useful and reliable than segmentation by firm size and e-government solutions.109 • E-Government solutions, on a yes/no basis. Also in this respect, the BRE observed ex post that segmentation by e-government solution did not prove to be particularly useful within the measurement exercise.110 • Other criteria: in its SCM manual, the BRE had accounted for the possibility of using other segmentation criteria, mostly related to existence of regional difference, outsourcing of certain activities, etc. 188.8.131.52 Population, rate, frequency The UK measurement initially followed the prototype SCM in setting criteria for identifying population, rate and frequency for each IO. In line with the prototype SCM, the UK model assumes 100% compliance with IOs, without accounting for instances of non-compliance. The BRE provided tables and statistics from the Office of National 108 KPMG used also the Nano category (0 employees) for the HMRC measurement exercise. 109This activity was carried out for each IO or data requirement (DR), but many IOs or DRs were found to apply across sectors, or to groups of organisations that did not correspond to SIC codes. This created some problems as SIC codes were found to provide an insufficient basis for defining the relevant population. 110Face-to-face interviews included questions on the use of e-government solutions – which, as a result, were accounted for at a later stage, after the segmentation and the identification of the relevant population. – 89 – Statistics and the Small Business Service. In addition, data from departments’ impact assessments and other materials were used in an initial screening. In practice, however, the methods used by the consultant were more sophisticated. For each IO/DR, the consultant identified an appropriate information metric, in order to describe the quantity based on three main categories: a) sector-based metrics; b) third party metrics; and c) number of transactions. The departments cooperated with the consultant in applying such metrics to the identified population, when available data provided at the outset were insufficient to achieve a complete picture. In most cases, 16 bands were used to segment the population, ranging from Band 1 (0-100) to Band 16 (>500,000). However, in practice this step has proved to be probably the most difficult in the whole implementation process: as reported by the consultant, this work has highlighted “extensive gaps in knowledge about the number of organisations that are affected by an IO/DR, the frequency with which they are affected and the extent of compliance”. This also led to an extensive use of data provided by departments, which in turn reported that estimates were very hard to reach in many instances. 184.108.40.206 Demarcation Unlike what happens, for example, in the Dutch implementation of the SCM, demarcation in the UK is not carried out on the basis of a default rule (e.g., a fifty- fifty split). To the contrary, departments reach an agreement whenever an overlap requiring demarcation occurs, and are required to show an audit trail confirming the negotiation process and the agreement reached. If an agreement cannot be reached, the BRE arbitrates. In practice, in the UK a default outcome was sought, by allocating IOs arising from a given regulatory initiative to the department that undertook such initiative. In the UK, the process was more difficult as the possibility of double counting also had to be verified with respect to the two separate measurements carried out, the HMRC and the Financial Services Authority. 220.127.116.11 Data sources and collection methods In the UK, a number of different methods were used to collect data needed for the baseline measurement. In particular, the measurement required 8,476 telephone interviews, 408 face-to-face interviews, 198 Expert Panels and 15,558 expert assessments. The percentage breakdown of measurement methods is illustrated in Table 25 below. As illustrated in the table, the type of measurement method used was made strongly dependent on the expected magnitude of the IO/DR to be measured. As a result, direct assessment was used for 55.7% of total IO/DRs, which however accounted for only 7.4% of overall estimated costs. To the contrary, over 17% of IO/DRs were measured with some contribution from business interviews, accounting for almost 80% of total estimated costs. 26.1% of the IO/DRs were measured only through Expert Panels, accounting for 13.9% of total costs. – 90 – Finally, as regards the interview process, a very detailed interview guide was provided by the consultant and is reported in its Technical Summary.111 Table 25 – Measurement methods Source: PwC, Technical Summary 18.104.22.168 The ‘normally efficient business’ As already recalled in Section 1, the application of relevant cost and time parameters to the activities to be measured necessarily entails the prior definition, for each segment, of the ‘normally efficient business’. This step is particularly critical, in that it entails in most cases a qualitative assessment by the consultant. In the UK, the BRE initially recommended the same approach adopted in the prototype SCM, which entail the identification of the media firm in the segment and the elimination of outliers in observing the reported data. However, in the practical implementation of the model in the UK, the issue of how to define the typical business had to be addressed in a number of different ways, depending on the type of data capture tool used. In particular, where expert panel were involved in the assessment, an anonymous ‘straw man’ costing (i.e. an estimate for a normally efficient person carrying out the activity within a normally efficient business) was usually completed by one participant prior to the Panel meeting, leading to a final decision of what a suitable estimate would be. Interestingly, the use of expert panels was found to be particularly useful. 22.214.171.124 Cost parameters and overhead The application of cost parameters to the collected data followed the general scheme provided by the prototype SCM, which includes internal costs, external costs and acquisitions: • The assessment of internal costs was carried out based on the Annual Survey of Hours and Earnings published by the Office of National Statistics. Nine staff types were identified, and for each of them a median hourly wage rate was applied.112 111 Quote, Appendixes 1, 2. – 91 – • The overhead applied to time-based activities was 30%.113 • External costs were mostly assessed on the basis of reported cash spend by respondents to business interviews. 126.96.36.199 Extrapolation, reporting and database At the end of the measurement exercise, the consultant extrapolated the data collected to national level, by programming into ABR.net, a bespoke project management information system used as a repository of all data used in the ABME. The actual content and presentation of the final report is discussed with the BRE and the department involved, and includes a list of the ‘top x’ (normally, 10) most burdensome regulations. The classification of regulatory origin maintains the differentiation between international, EU-Directive and EU-regulation, and this would allow for identifying the most burdensome IO/DRs generated by pieces of EU legislation. Each department must file its own final report.114 The publication of Final Reports is expected by October-November 2006. The reporting template includes a summary of the measurement, which illustrates key results, total costs and costs per affected business. In addition, a summary of the origin of regulation has to be provided. The main sections of the final report are the following: • Introduction, containing a background section, a summary of departmental regulatory areas and units/agencies, a brief outline of the methodology and of key stages of the process, a description of the organisation and process of the measurement, and a statement on the quality/properties of the measurement undertaken; • An overview of regulations affecting businesses, with summary statistics of regulation included in the measurement, a breakdown of number of regulations, IOs and DRs for each unit composing the department, a description of main characteristics of the regulated businesses and a methodological note. • An overview of administrative costs broken down for each unit composing the department, with an indication of total costs per unit, percentage on total costs and costs per affected business. Results should also be presented by firm size and – where possible – industry sector. • A breakdown of administrative costs by regulatory origin, including a breakdown of categories A and B by international, EU-Directive and EU-regulation origin. 112A standard value of time was derived for each of the categories based on the median hourly wage rate (excluding overtime) across the United Kingdom in 2005 by matching the ASHE occupational groups to those used in the data gathering and then calculating the weighted average median wage using the number of jobs in each occupational group as the weight. 113 See Annex 3 of the UK SCM Manual for a detailed breakdown of elements included in the overhead. The only example available is the Final Report of the Forestry Commission, completed and signed of in June 2006 114 and published in August 2006. – 92 – • A section illustrating ‘top x’ (often, top 10) most burdensome regulations by total administrative costs as well as by administrative costs per affected business, broken down by regulatory origin (ABC). • A list of administrative costs by activity, based on the list of standard activities illustrated above, in Table 2. • A section containing detailed data on administrative costs by departmental unit, including a table listing IOs and DRs and broken down – where possible – by firm size and e-government solutions for the top x most burdensome regulations. In addition, the report must contain a summary by standard activity type in which the department clarifies the relative weight of form-filling for each activity. • A summary section at the end. • Annexes containing a step-by-step description of the measurement process, an illustration of cases in which demarcation was necessary, an example interview guide, a table on annual administrative costs by IO with details on the population and breakdown of interview types. Standardized templates are offered by the BRE for all the content of the final reporting. 188.8.131.52 Targets The UK government committed to setting targets for the reduction of administrative burdens by the pre-budget report 2006. Outcomes, simplification plans and targets will be set for each department, and will be formally submitted to the BRC and to the Panel for Regulatory Accountability for scrutiny. In terms of overall target-setting, the Prime Minister has announced a 25% reduction target for three departments (DTI, DEFRA and HSE) after the publication of draft simplification plans in November 2005.115 Subsequently, draft simplification plans were also published for the Department of Health and the FSA. In any case, targets by Department will eventually be set, and will be all around 25%. If a differentiation will be made, it will entail only minor differences between departments. 2.3.3 Results The overall results of the UK measurement will be made available by the end of 2006. However, the BRE could already make available some of the most important findings for categories A and B, which are still to be considered as confidential information and, accordingly, are not included in this version of the Final Report. The sophisticated model used in the UK allows for a double representation of the results – in terms of 115 DTI has already taken action by abolishing the requirement for 400,000 small businesses to re-apply for small business rate relief every year. A detailed plan was published by DTI to deliver annual savings of approximately £220 million yearly in the period 2005-2010. HSE has committed to extend risk-based inspection and enforcement practices across its operations and to cut the burden of form-filling by 25%. In most cases, the sectoral targets identified so far are still ‘arbitrary’. Furthermore, it is not clear whether the 25% is calculated on the baseline or on other measures, or on a more specific feature (e.g. form-filling burdens). – 93 – administrative costs (total costs) and in terms of so-called “net costs” (i.e., administrative burdens), whereas the latter do not include costs related to activities that businesses would have carried out even absent regulation. From this standpoint, the UK measurement appears very flexible and easy to compare with both country measurement based on administrative costs, and measurements based on administrative burdens. The results that will be published at the end of 2005 are only part of the evidence gathered so far by the UK government in its attempt to implement the SCM and reduce administrative burdens. As already mentioned, a separate measurement – also based on the SCM – was carried out for two departments, the Tax and Customs department (HMRC) and the Financial Services Authority. In the HMRC exercise, a consortium led by KPMG calculated that the administrative burden of UK tax regulation equalled £5.1 billion at 31 May 2005.116 EU legislation accounted for 34% of overall administrative burdens (15% in category A-EU- Regulation, 16% in category A-EU-Directive and 2% in category B-EU-Directive). Three of the most burdensome IOs were derived from EU legislation: the issuing of VAT invoices (9% of total), the submission of VAT return quarterly (6%) and the issuing of customs declarations (5%). The first two, of course, relate to an EU Directive, whereas the third arises from an EU Regulation. As regards the study conducted by Real Assurance Ltd. For the Financial Services Authority, it covered 233 FSA rules, and some 340 firms (broken down into small, medium and large in line with the taxonomy used by the SBS) across financial services. According to the study, the total administrative burden imposed by the 158 rules identified as imposing such costs is around £600m per annum, approximately 0.5 per cent of industry costs. The highest burdens fall into two general categories: • high costs incurred by a few very large firms, such as costs imposed by rules on money laundering prevention, which account for around 40 per cent of the total estimated costs; • rules which apply to almost all the 25,000 firms regulated by the FSA. According to the study, the top 20 administrative burdens cover rules that account for more than 85% of the total administrative costs. In general, however, this study exhibited a lower degree of detail if compared to the general measurement and the HMRC measurement. For example, the final report does not address the issue of the origin of administrative burdens. 2.3.4 Most burdensome EU Regulations The observation of results of the UK baseline measurement allowed for the identification of the most burdensome EU regulations. This measurement can be used as a benchmark for other countries, which in some cases have indicated the same 116 The Project took place between September 2005 and March 2006. – 94 – regulations as generating significant administrative costs for private enterprises. The estimated total administrative cost generated by EU regulations is £622,365,617.62, which is approximately 6.2% of total administrative costs in Categories A and B and 8.2% of net costs. This figure also equals 7% of total administrative costs generated in the UK by EU legislation, and 8.3% of net costs. It must be warned, however, that these figures do not include regulations analysed in separate measurement exercises, such as those carried out for the Financial Services Authority and the HMRC. 2.3.5 Summary and comments The UK implementation of the SCM is particularly ambitious and sophisticated both from the standpoint of the methodology adopted, and because the SCM is integrated with the established better regulation tools in use at the Cabinet Office and in other levels of government. This level of accuracy is also reflected in the overall cost of the exercise: the identification of the pieces of legislation to be measured and the baseline measurement are expected to reach a total cost of approximately €26 million.117 According to the BRTF (now BRC), the organisational costs for the reduction work will be around €5.9 million (£4 million) per year in the UK, compared with €5 million in the Netherlands.118 In this regard, the most difficult problems emerged during the baseline measurement were the volume of data to be collected; the determination of the relevant population affected – where the use of expert panels proved useful and not costly – and most importantly prioritisation issues. The application of a proportionality criterion in the UK led more than half of the IO/DRs to be measured through direct assessment. In addition, the methodology used in the UK to measure administrative costs ex post is slightly different from the one adopted for ex ante measurements. In particular, as is understandable, RIAs normally consider only administrative burdens, not administrative costs: this means that, unlike the SCM, a RIA will not cover the cost of activities that businesses perform anyway, regardless of the existence of a legal obligation. This has to be carefully taken into account when updating the baseline with new pieces of legislation for which a RIA has been undertaken. This potential problem is even more relevant as the BRE favoured an approach aimed at creating a ‘live’ baseline, constantly updated with new RIAs, instead of launching periodical baseline measurements, for example, every three years.119 As observed by the BRE, a second round of baseline measurement (such as that being launched in 2007 in the Netherlands) would always lead to results that are impossible to compare with the results of the previous measurement, due to approximation and extensive 117 Quote from SCM Questions. The UK have put the one-off measurement cost at £11.4 million. 118 Like in the Netherlands, also in the UK 3-5 persons per department are involved in the measurement process. In the UK, since August 2000, RIAs must include details of “implementation costs” of new pieces of legislation. 119 Implementation costs are defined very similarly to administrative costs (or AB?) in the SCM. – 95 – sampling. On the other hand, it is recommended that individual regulations are reviewed every three years. In summary, it can be concluded that the UK implementation of the SCM is of particular interest for the development of the EU common methodology, although some of its features would be hard to replicate at the EU level and its cost is probably excessive in this respect. The possibility of tracing back the original piece of legislation from the ABR.net tool would enable full communication between the EU and the BRE in apportioning responsibility for the reduction of administrative burdens. However, gaps still exist in the database which can hamper attempts to trace back the original piece of EU legislation. With this caveat, when all the results of the baseline measurement will be available, the database will become a precious tool also for the forthcoming EU Action Programme. In this respect, it is worth mentioning – although shortly – that the Davidson review of over-implementation of EU legislation published the results of its call for evidence on 19 July 2006. The call for evidence ran from 3 March to 25 May 2006, and generated just over 160 written responses from representatives of many industry sectors. The final report of the Davidson Review will be published with recommendations by the end of 2006. – 96 – 2.4 The Czech Republic The measurement exercise in the Czech Republic started in March 2005 with a government resolution adopted in April 2005: departments/ministries, and administrative authorities were asked to measure administrative burdens in their areas, supported by the Better Regulation Unit. The measurement was basically oriented at the Dutch Standard Cost Model; however, a decentralised approach was pursued: i.e., departments/ministries and administrative authorities involved were given considerable leeway concerning the individual steps of the measurement exercise. 2.4.1 Methodology The baseline measurement is confined to IOs imposed on private businesses, as these are easy to define and to identify, and because the government generally aims at improving the business environment for the private sector. The measurement covers IOs not only to public authorities, but also to employees and customers. The main reason for such choice is that it can be safely assumed that IOs to third parties make up for a significant share in the overall administrative burdens borne by private businesses, and because it would be difficult to disentangle administrative burdens according to the individual parties which are entitled to the information that businesses are obliged to provide. 2.4.2 Content The areas covered by the measurement exercise were Finance, Education and Sports, Culture, Labour and Social Affairs, Health, Justice, Interior Affairs, Industry and Trade, Regional development, Agriculture, Transport, Environment, Informatics, Statistics, Surveying, Mapping and Cadastre, Mining, Industrial Property, Protection of Competition, Nuclear Safety, Securities, National Security, Energy, Telecommunications, Radio and Television Broadcasting. The measurement exercise showed that only 3 areas did not create any administrative burdens on businesses (Foreign Affairs, Defence and State Material Reserves), because these areas do not include regulations generating IOs for private businesses. This assumption was substantiated by discussions within the interministerial group built for preparing the measurement exercise. No minimum threshold (e.g. 100 hours per year) was applied for the baseline measurement, nor is one planned for the future: the measurement is intended to capture the administrative burdens as comprehensively as possible and to provide the basis for a complete measurement of the long-term development of administrative burdens, which would be precluded by excluding certain IOs due to a minimum threshold. – 97 – 2.4.3 Procedure There is a central coordination unit for the measurement of administrative burdens in the Czech Republic: the Department of Regulatory Reform and Central State Administration Reform – Better Regulation Unit, which is an Office of the Government. In general, a rather decentralised approach to measure administrative burdens was followed in the Czech Republic; this can be considered as the most significant deviation from the SCM (which represents the methodological foundation of the Czech measurement exercise). Apart from that, the measurement exercise was closely oriented at the manuals devised by “pioneer countries” (particularly the Dutch ones). In the meantime, the Better Regulation Unit has adjusted the methodology, taking into account the experience made during the measurement exercise.120 Moreover, as the introduction of a mandatory impact assessment of new legislation is planned for 2007 in the Czech Republic, the methodology had to be adjusted accordingly to allow also for ex ante assessments. Hereby, no “one in, one out” approach will be pursued, considering the Czech Republic’s specific situation as a new European member state obliged to implement the whole European acquis communautaire. Therefore, this specific situation turned the intended ex-ante impact assessment and the ex-post measurement of administrative burden into independent projects. In principle, an identical methodology will be used for ex ante assessment and ex post measurement. However, the Better Regulation Unit is waiting for an adoption of the government resolution which is not released yet concerning these new aspects of the methodology. Furthermore, the implementation of a detailed survey enumerating the concrete legislation proposed for the purpose of the reduction of administrative burdens is considered. The survey will be carried out by 5-6 ministries – also the Better Regulation Unit will participate – that are responsible for 90% of total administrative burdens. Administrative burdens are broken down by ministries/departments and administrative authorities. As already recalled, ministries/departments and administrative authorities were quite independent in conducting the concrete measurement procedure. This decentralised approach was chosen mainly for two reasons: first, to account for existing differences across ministries/departments and administrative authorities and their respective areas of competence; secondly, due to the very limited staff and resources of the Better Regulation Unit, which made it necessary to delegate the lion’s share of the measurement exercise to the individual public authorities under scrutiny. It must be seen as a considerable disadvantage of this decentralised approach that the Better Regulation Unit does not have central evidence concerning the concrete measurement procedure followed by the individual ministries/departments and administrative authorities. The measurement exercise is not based on a standardised list of IOs, as it was deemed impossible to come up with a complete and comprehensive list. The methodological instructions and information provided to the ministries/departments and administrative authorities in charge of the measurement featured, however, a list with 120 The latest version of the methodology is currently being translated into English. – 98 – examples of IO. However, the new version of the methodology will contain a more detailed list. Moreover, ministries/departments and administrative authorities were offered the possibility to contact the Better Regulation Unit to clarify whether certain rules and activities represented IOs and should therefore be included in the measurement. Information obligations are classified as EU or non-EU related according to the regulatory origin, which follows the standard ABC classification. As regards demarcation issues, when the same regulatory requirement can be attributed to more than one ministry/department, the information obligations and costs are equally split between them, like in the Dutch and Danish models (and unlike what occurs in the UK). Businesses included in the measurement exercise were selected based on various criteria; as already mentioned, there were no central binding requirements or specifications. In this respect, the recommendation provided by the Better Regulation Unit was to segment businesses according to form size (in terms of number of employees)121: • entrepreneurs (no employees) • small businesses (up to 10 employees) • medium businesses (10 to 250 employees) • large businesses (over 250 employees) All firm sizes should be represented in the sample, provided that they are affected by the information obligation. Most ministries did not exploit the possibility to use additional criteria, e.g. turnover. If ministries presumed that the administrative burden of a certain IO was independent of size, no differentiation was made. The recommended minimum size of the sample was 5. The ministries/departments and administrative authorities used various data sources in the selection process. The main data sources used were (sorted by importance): statistical data from the Czech Statistical Office; data obtained from businesses; own statistical data of ministries. The Better Regulation Unit recommended integrating national statistics with EU statistics; however, the latter were only seldom used. The selection criteria used cannot be specified, as the selection process was in the discretion of the individual ministries/departments/administrative authorities. However, generally the selection criteria applied did not differ broadly across different sectors. In this respect, it is also of interest that some of the individual ministries/departments/administrative authorities, due to the nature of their activities, have easier access to private businesses than others, which might also have influenced the selection process (see also below). 121Note that this classification is different from the one used in the UK, as recommended by the Small Business Service. See above, Section 2.3. – 99 – Focus groups were the most frequently used data collection method, followed by telephone interviews. Also expert assessments and face-to-face interviews played a role. The predominant role of telephone interviews is due to cost considerations and because conducting interviews in many cases was the most simple and convenient method. Some ministries could rely on existing contacts with private businesses (e.g. the Ministry of Industry and Trade), while 2 or 3 ministries (e.g. the Ministry of Health) had to make use of the support of the Chamber of Commerce to establish contacts to private businesses (for which service the Chamber of Commerce had to be reimbursed). No uniform interview guide was used: thus, the interviews conducted were not based on standardised questions. According to the information of the Better Regulation Unit, several ministries/departments and administrative authorities just sent some tables to the businesses asking them to fill in the data, without really proceeding to interview them, and sometimes without providing further instructions and explanations. Where expert assessments were applied, they were conducted by internal experts at the ministries/departments or administrative authorities. Workshops, field trials, business panels and consultancy studies (no external consultants were involved) were not used. According to the future methodology, which will also cover ex ante assessments, expert assessments will be of greater importance than they were for the baseline measurement. The question of the overhead rate was not specified by the Better Regulation Unit in the beginning of the measurement process, as it was not addressed in the first versions of the Dutch manuals on which the Czech measurement exercise was based. As overhead was included in later versions of the manuals, the Better Regulation Unit only recommended a percentage of 25% for all ministries/departments and administrative authorities; the calculation of the overhead percentage, however, was left to the individual ministries/departments and administrative authorities. The measurement exercise includes all voluntary IOs – i.e. those that would be complied with even if compulsory regulation would be removed – as well as voluntary obligations (e.g., when applying for subsidies). However, only IOs based on legal provisions were measured (thus, for example, codes of conducts concluded on a voluntary basis were excluded). Normally efficient businesses, i.e. businesses that handle their tasks in a normal manner, were identified by the individual ministries/departments/administrative authorities in charge of the measurement. No central standard for the normally efficient business was set; the Better Regulation Unit only provides some general recommendations how to identify normally efficient businesses (according to the international manuals). There is no central evidence available how normally efficient businesses are identified. Businesses are segmented by firm size and industry sectors. The segmentation according to industry sectors is based on the Czech industry classification. In principle it would be possible to use the NACE classification in the future. – 100 – No standardised measures of time/cost for certain administrative activities are used. The determination of internal and external prices and tariffs was left to the individual ministries/departments/administrative authorities, who used their own expertise or asked businesses to provide the necessary information on external and internal costs. As a result, no standardised (internal and external) tariffs were applied. Table 26 provides an overview over the distribution of tasks and responsibilities connected with the 15 steps of the standard cost model. Table 26 – Who Does What in the 15 Steps of the SCM – Czech Republic Consultant(s) Coordinating Ministry(ies) Central Phase Step Unit Description Identification of business-related regulation to be included in Phase 0 0 X X the analysis Identification of information obligations, data requirements 1 X X and administrative activities and classification by origin 2 Identification and demarcation of related regulation X X 3 Classification of information obligations by type 4 Identification of relevant business segments X Phase 1 5 Identification of population, rate and frequency X 6 Business interviews v. expert assessment X 7 Identification of relevant cost parameters X 8 Preparation of interview guide X 9 Expert review of steps 1-8 10 Selection of normally efficient business X 11 Business interviews X Phase 2 Completion and standardisation of time and resource estimates 12 X for each segment by activity 13 Expert review of steps 10-12 14 Extrapolation of validated data to national level X Phase 3 15 Reporting and transfer to database X No specific (and additional) budget was allocated to the measurement exercise, which therefore was carried out without the support of external consultants. The measurement was coordinated by the Better Regulation Unit within its regular duties, – 101 – which do not only consist of measuring administrative burdens. The total budget allocated to the Better Regulation Unit cannot be split according to its individual tasks, as the Czech government does not apply project-oriented budgeting. The ministries/departments/administrative authorities involved, which did most of the measurement as part of their regular activities, did not receive any additional resources. Altogether, the measurement exercise started in March, 2005, and was completed at the end of December, 2005.122 2.4.4 Targets Before starting the measurement exercise, the Czech government set an overall reduction target of 20% of the total administrative burdens, which should be reached by the end of 2010. An overall target was chosen, in line with the positive experience of other countries, mainly as an incentive for reduction. The overall reduction target levels fixed by the other countries served as an orientation for the Czech reduction target, which, however, decided to opt for a slightly less ambitious percentage. The possibility of adjusting the target after completing the measurement exercise to its results was left open; however, the probability that the new government in office since mid-2006 will change the target is rather low. In selecting the concrete timing of the overall reduction target, account was taken both of the experience of other countries and of national peculiarities (particularly the available resources and the upcoming elections). No sector-specific reduction targets were set, as this was not deemed useful before knowing the results of the measurement exercise. Currently the most important precondition to take further action to reduce administrative burdens is the passing of the still pending government resolution concerning • the implementation of mandatory ex-ante administrative burden assessment and the approach “one in, one out” in case of laying IOs of category C (IOs are set entirely by the Czech competent bodies) • adoption of the new version of the methodology and • implementation of a detailed survey enumerating the concrete legislation proposed for the purpose of the reduction of administrative burden. The survey will be carried out by 5-6 ministries – the Better Regulation Unit will also participate – that are responsible for 90% of total administrative burdens. If the government adopts the new methodology by the end of 2006, the target year 2010 can be maintained; otherwise it would have to be adjusted accordingly.123 122Originally the measurement exercise was scheduled for 6 months, i.e. it was to be completed by the end of September, 2005. As several public authorities involved could not finish the measurement on time, however, the deadline was extended to the end of December, 2005. 123 Currently it is uncertain when the resolution will be passed, due to the instability of the overall political situation. – 102 – Finally, like the Netherlands, the Czech Republic opted for a model entailing successive rounds of baseline measurement. As a result, in order to measure progress in the reduction of administrative burdens, a second measurement exercise is planned for 2010. 2.4.5 Results The overall administrative burdens measured in the Czech Republic amount to approximately €3 billion, which corresponds to 2.9% of GDP and € 300 per capita respectively. 2.4.6 Most burdensome Directives In the Czech baseline measurement, a number of EU directives were identified as particularly burdensome for private businesses. These are reported below, in Table 27 As is easily seen in the table, it was not possible to identify most affected sectors, and only in some cases an estimated figure for administrative costs generated by the individual piece of EU legislation could be reached. – 103 – Table 27 – Most burdensome EU directives in the Czech Republic 2.4.7 Summary – Lessons The implementation of the SCM in the Czech Republic was developed and carried out with a very limited budget, no consultants and a high degree of flexibility, especially as far as the discretion of ministries/departments and administrative authorities is concerned. The less ambitious target (20%) was set as a political decision, mostly for incentivising agencies to achieve significant reductions over the set timeframe. The Czech model could be improved in a number of ways, which includes the following: • Strengthen the central collection of data and evidence (by the Better Regulation Unit) on all stages and results of the measurement exercise; – 104 – • Adopt a standardised list of IOs and associated DR/activities; • Specify criteria for selecting businesses and use harmonised criteria for segmentation and sampling; • Use a uniform interview guide with standardised questions; • Specify the procedure to identify the ‘normally efficient businesses’; • Use standardised measures of time and cost (external and internal tariffs) for certain administrative activities; • Allocate a specific and additional budget to the measurement exercise, also to be able to pay external consultants; • Appoint one permanent contact person in each department as the link to the Better Regulation Unit; • offer more training for people involved in departments. – 105 – 3 CROSS-COUNTRY COMPARISON AND MAIN FINDINGS The four SCM variants illustrated in the previous section can be compared from a methodological, procedural and organisational perspective, as well as from the standpoint of results reached. In this section, we describe the main similarities and differences between the selected models. In addition, we highlight the most burdensome pieces of EU regulation as emerged from the results of the baseline measurements in the four countries analysed. Finally, we identify a list of priority areas for simplifications. 3.1 Cross-country comparison The implementation of the SCM tool in the four countries analysed exhibits significant similarities, but also important differences as regards the aims/purposes of the measurement, the scope and frequency of the measurement, the definition of administrative burdens, and the organisational arrangements adopted. Some of these differences can be considered as relevant in light of the upcoming EU-wide measurement of administrative costs. Table 29 below compares some of the key features of the four models and of the EU SCM as described above, at Section 1.7. The most relevant differences that emerge from our case studies are the following: • Scope of measurement: some countries, such as the Czech Republic and Denmark, only calculate administrative costs faced by private enterprises, whereas the UK calculate costs for private and semi-private businesses, and the Netherlands also include citizens. The EU model is the most ambitious, in that it also includes public administrations. • Targets: most countries have selected a 25% ‘political’ target reduction of administrative costs. An exception is the Czech Republic, where the target selected is 20%; in addition, some countries – such as the Netherlands and, to a lesser extent, the UK – decided to set (at a later stage) specific targets for ministries/departments, whereas others – like Denmark and the Czech Republic – have not set different targets for specific ministries. • One-off costs: most countries – with the exception of the Czech Republic – do not include one-off costs in the baseline measurement. However, in the UK model one- off costs are included in the overhead, and the UK and Denmark include one-off costs in the ex-ante measurement carried out within RIAs. The proposed EU model specifies that such costs, when non-marginal, ‘may be taken into account’. This was also the approach chosen for the Dutch baseline measurement.124 • Obligations set by optional schemes: these information obligations are commonly referred to as ‘voluntary obligations’. In Denmark and the UK, these IOs are taken 124 See Manual, December 2003, p.15. – 106 – into account, if they are expected to arise from a quasi-regulation. In the UK, the baseline includes also codes of conduct and guidance documents with government backing, which cannot be considered as having binding force. • Distinction between ‘pure’ obligation and good practice written in the law: ‘Pure’ obligation refers to what one would stop doing if the legal obligation was removed. The EU model assesses all legal obligations even when the latter correspond to good practices. In other models, it is possible to assess only pure obligations. • Overhead: in the UK, a 30% rate was chosen, compared to the 25% rate used in the Netherlands and in Denmark. In the Czech Republic, the overhead percentage has been set by particular ministries. No standard overhead percentage for all ministries and administrative authorities is intended. • Demarcation: most countries use a default 50:50 split whenever an IO/DR is found to fall under the competency of more than one department. In the UK, the need for demarcation was foreseen, but no 50:50 split rule was chosen. The most appropriate split is therefore set through negotiation between departments, and normally the department that has taken the regulatory initiative keeps ownership of the corresponding obligations. • Segmentation: in practice, segmentation criteria differed noticeably. In the UK, industry sector and type of activity proved more useful than firm size (except for the tax measurement undertaken separately). In the Netherlands, Denmark and the Czech Republic firm size was a guiding principle. • Accuracy and costs: the Danish and UK models seem likely to reach a higher level of accuracy in the estimates, also due to the intensive involvement of consultants, the more detailed breakdown of IOs into data requirements (Denmark) and the further breakdown of the classification of origin (Denmark and UK). The UK database seems the most suited to retrieve original pieces of EU legislation, although with some gaps. The greater accuracy, however, is reflected also in the cost of the measurement, especially for the UK. • Organisational patterns: as shown below, in Table 28, the four national models exhibit some differences as regards the role played by ministries, central coordinating units and consultants. The Dutch and Czech models rely on ministries much more than the Danish and UK models. In the UK, ministries/departments are involved only in the initial step of the measurement, which is then carried out mostly by the consultant with the supervision of the BRE. In the Netherlands, IPAL is involved mostly in the review of activities performed by ministries and consultants, and in the final stages of the measurement. In the Netherlands, consultants are not involved in the final reporting and transfer to database (step 15), whereas in Denmark and the UK the setting up of a database was achieved with a strong involvement of consultants. – 107 – Table 28 – Who does what – cross-country comparison In summary, the analysis of the four case studies allows us to draw two main conclusions. On the one hand, the adoption of a common theoretical framework – provided by the SCM tool – will certainly represent a significant advantage for the implementation of a multi-level measurement model coordinated by the European Commission. Had existing models obeyed to differing theoretical assumptions and methodologies, such an exercise would certainly have proven impossible from the outset. At the same time, it must also be observed that the differences in the four models analysed makes cross-country comparison a rather challenging exercise. The International SCM Network has recently issued guidance on how to conduct cross- country comparisons, and specified that, before conducting a benchmarking exercise, the following minimum points must be agreed upon ex ante: • Definitions: countries must use the same definition of IO, government regulation, business etc.; • The level of detail in the measurement. • Collected data: it must be decided which data will be collected and how they will be reported. For example; are the businesses segmented by size (small, medium and large firms) or not? • How is the overhead calculated? • What is the approach to activities that a business may choose to continue in the absence of regulation (normal business costs)? • How are one-off costs dealt with? – 108 – Accordingly, even if only these points are considered, any real benchmarking between the four countries analysed in section 2 above would seem hardly feasible. As already recalled, significant differences exist in most of the abovementioned points. A first attempt to compare national models was made within the “International comparison of measurements of administrative burdens related to VAT in The Netherlands, Denmark, Norway and Sweden” carried out in 2004-2005.125 There, differences in the methodology adopted by the four member states compared created substantial challenges in ensuring actual comparability of results. The final report states clearly that: “there are some differences in how countries have carried out the measurements. Figures that would appear to be simply comparable are found to show substantial differences … In some cases, the countries have measured information obligations that initially have appeared to be the same, but after a deeper analysis it has shown that the countries have made various limitations within the legislation. In some cases it has been possible to make some adjustments and then compare the figures. In other cases the differences have been too big for a correct comparison” To be sure, not all the problems encountered in the VAT international comparison are relevant for the creation of an EU SCM and the implementation of the upcoming Action Programme. However, some of these differences must be considered as important in this respect. For example, particular problems emerged as regards the classification of origin, which had to be made at IO level. It emerged that the same IOs had been classified very differently by the benchmarked member states, and that the resulting weight of category A, B and C burdens differed widely.126 This result also suggests that, in setting up an EU common methodology, involvement of the Commission in the classification of regulatory origin should be ensured. 125 See www.administrative-burdens.com. 126 See Table 12 in the Final Report of the International Comparison on VAT. – 109 – Table 29 - Comparison of SCM variants Dutch SCM Danish SCM UK SCM Czech SCM EU SCM Assessment of the costs of Assessment of the costs of Assessment of the costs of Assessment of the costs of Assessment of the costs of Aims/purposes/ administrative obligations imposed administrative obligations imposed administrative obligations imposed administrative obligations imposed administrative obligations imposed targets on private and semi-private on enterprises by future or existing on private and semi-private on enterprises on enterprises, the voluntary enterprises and citizens, business related Danish laws, enterprises by future or existing sector, public authorities and distinguishing between national and distinguishing between national business related UK laws, including citizens, distinguishing between non-national origins (ABC and non-national origins (detailed also guidance documents and codes international, EU, national and classification) ABC classification) of practice with government regional origins backing, distinguishing between national and non-national origins (detailed ABC classification) Both microeconomic and Both microeconomic and Both microeconomic and Both microeconomic and Both microeconomic and macroeconomic use macroeconomic use macroeconomic use macroeconomic use envisaged macroeconomic use 25% overall reduction by 2007, with 25% overall reduction by 2010. No No overall target set. Sectoral 20% overall reduction by 2010. No n.a. differing targets for each ministry sectoral targets targets under discussion sectoral targets “costs imposed on businesses, when The costs regarding the the costs of administrative activities The cost of compliance with “costs incurred by enterprises, the Definition of complying with information administrative activities that that businesses are required to information requirements ensuing voluntary sector, public authorities administrative obligations stemming from businesses have to carry out in conduct in order to comply with the from legal regulations is referred to and citizens in meeting legal government regulation.” order to comply with the information obligations that are as administrative burden. obligations to provide information cost information obligations that are imposed through central on their action or production, either “AB Citizen comprises the costs imposed through official regulations government regulation. to public authorities or to private incurred by citizens in complying parties.” with information obligations ensuing from government regulations. It includes both compliance with obligations and the exercise of rights.” One-off costs not taken into account One-off costs not measured in The baseline measurement does not One-off costs – which a business is One off costs may be taken into baseline, but in ex-ante cover one-off costs. However, such required to expend to comply with account measurement and in yearly update costs are taken into account in the the information obligation, should RIAs. also be evenly spread over the period being analysed Actions that would have been Actions that would have been Actions that would have been Actions that would have been Actions that would have been undertaken even absent legislation undertaken even absent legislation undertaken even absent legislation undertaken even absent legislation undertaken even absent legislation are taken into account are taken into account are taken into account are taken into account are taken into account – 110 – Σ=Ρ•Q Σ=Ρ•Q Σ=Ρ•Q Σ=Ρ•Q Σ=Ρ•Q Core equation Price (P) = tariff x time Price (P) = tariff x time Price (P) = tariff x time Price (P) = tariff x time Price (P) = tariff x time Quantity (Q) = n. of businesses x Quantity (Q) = n. of businesses x Quantity (Q) = n. of businesses x Quantity (Q) = n. of businesses x Quantity (Q) = n. of businesses x frequency frequency frequency frequency frequency Focus on labour costs, with Focus on labour costs Focus on labour costs Focus on labour costs Where appropriate, types of costs extension to citizens other than wages and overheads will be taken into account The SCM is applied to all regulatory The SCM is only applied to rules The SCM is applied to rules that The SCM is applied to all regulatory Ex ante application only to Scope and proposals and acts in force. that cover active private Danish cover active private and semi- proposals and acts in force that proposals imposing major frequency businesses (all obligatory and private UK businesses (all affect private businesses. administrative requirements and/or certain types of voluntary rules). obligatory and voluntary rules). acts identified as particularly burdensome by end users Applied to all administrative actions Applied to all information Applied to all information Applied to all information imposed by a piece of legislation, requirements imposed by a piece of requirements imposed by a piece of requirements imposed by a piece of Use of thresholds to identify most except for exceptional (and legislation, with threshold of 100 legislation (no minimum thresholds) legislation (no minimum thresholds) onerous actions undefined) marginal costs hours administrative work per year IOs to third parties are measured IOs to third parties are measured IOs to third parties were measured IOs to public authorities, employees Review and timeline for ex ante in practice, and will likely be and customers assessment defined on a case-by- reported separately case basis Periodic review (4-5 years) Yearly update of the baseline Baseline is kept live through RIAs Ex-ante impact assessments planned Periodicity of cumulative/sectoral measurement with new / altered and reviews of individual in the next years (from 2007 on). ex post assessments yet to be regulation and ex-ante legislation, with no expected second Second baseline measurement in defined. measurements baseline measurement 2010 Different percentages set by Overhead 25% (but also higher percentages are particular ministries. Better 25% 30% n.a. being used) Regulation Unit only recommended 25% Yes, in line with international Yes, in line with international Yes, in line with international No Yet to be defined Standardised manual manual manual, but with shorter list of lists administrative activities (7) – 111 – Accuracy sought mainly through Highest level of accuracy through High level of accuracy sought Lower accuracy due to lack of Use of range for administrative Expected level fieldwork and simulation subdivision into data requirements mainly through extensive fieldwork resources and absence of costs in ex ante impact assessments of accuracy and and simulation consultants to avoid spurious accuracy data sources Application of the proportionality principle to establish optimal accuracy levels. Use of national registers and Use of national registers and Use of national registers and Use of national registers and Use of a sample of member states statistics statistics statistics. Expert panels very statistics and extrapolation to EU level important during the measurements Main data collection methods were Main data collection methods were Main data collection methods were Focus groups were the most Estimates base on EU statistics, face-to-face and telephone face-to-face and telephone face-to-face and telephone frequently used data collection standardised ratios and data interviews, expert assessments and interviews, expert assessments and interviews, expert assessments and method, followed by telephone provided by the sampled member business panels focus groups business panels interviews. Also expert assessments states. and face-to-face interviews played a role. Classification Decision tree in 7 steps Decision tree in 7 steps Decision tree in 7 steps Decision tree in 7 steps Decision tree in 4 steps of origin Each public entity assesses its Ministries assess their regulatory Departments assess their regulatory Ministries assess their regulatory Application of the subsidiarity Division of regulatory proposals and sectoral legislation (with involvement of legislation (with involvement of legislationunder the supervision of principle: Commission assesses responsibility legislation (with involvement of consultants) under the supervision consultants) under the supervision the Department of Regulatory ‘upper bound costs’ and extrapolate consultants) under the supervision of the Division for Better Business of the Better Regulation Executive Reform and Central State national data to EU level of the IPAL and monitoring by Regulation (DBR) in the Danish (DBR). Consultants carried out most Administration Reform – Better Actal Division foe Better Business of the 15 steps. Regulation Unit (Part of the Office (The sample of) member states must Regulation under the Ministry of of the Government) provide data on their Economic and Business Affairs implementation of EU legislation. Need for interoperable databases Need to streamline communication between Commission services and responsible national authorities. – 112 – Methodological assumptions are Use of threshold – administrative EU legislation can be retrieved from Decentralised approach left too Inclusion of caveats clearly drawing Methodological clarified at the outset. activities requiring less than 1000 the database, but with some gaps. much discretion to departments, attention to the underlying features/caveats hours yearly are not measured. with negative impact on accuracy assumptions and their effect on the accuracy of the assessment. No standard interview guide was No standard interview guide was Analysis was “trial and error”, and No interview guide was used. used. used. excessively costly. The Dutch database does not allow Collection of qualitative information Use of expert assessment in line No consultant was hired for lack of for tracing back the individual piece is useful for identifying irritation with the proportionality principle, budget resources. of EU legislation burdens and areas for with potential shortcomings in simplification. Of course, it requires terms of accuracy No standardised list of IOs was interviews and focus groups, not used direct assessment. Methodology used in ex post measurement is slightly different No central standard for the from the one used for ex ante RIAs normally efficient business was set 3-5 dedicated and approx €300,000 Overall budget (for consultants) Identification of the pieces of Overall budget not available 14-40 hours of work over 4-24 weeks Workload and per ministry approx € 2 million, which includes legislation and baseline (decentralised approach, no project- per ex ante assessment cost pilot measurements, baseline and measurement are expected to cost orientated budgeting) the first yearly update of the around €26 million. Baseline Around 60 people involved in measurement. One ex post update- measurement alone will cost £11.4 public administrations (IPAL, Actal, measurement in a big area (e.g. million. Organisational costs for the ministries) annual accounts) cost about reduction work around €5.9 million €100.000 (£4 million) per year in the UK In 2002, overall cost for the baseline Equivalent of 6-7 fulltime staff in the 3-5 people employed for each Budget: no external consultants, 1,600 hours/week per year for the measurement was approximately €3 coordinating unit, equivalent of 1 department. most of the work done by ministries Commission’s central policy unit(s) million (done by consultants) halftime staff member per ministry for 8 months during the measurement Ex ante measurement is normally completed within one month at €10,000-20,000 per act. – 113 – 3.2 Priority areas for simplification The four case studies conducted within this Pilot Project allowed us to identify a list of policy areas that account for a substantial portion of overall administrative burdens imposed by legislation on businesses. This list is meant to include areas that should be awarded priority when approaching the issue of sectoral measurements. For the list to prove useful for such purpose, a number of caveats must be expressed at the outset. First, the results of the baseline measurements undertaken by the member states exhibit similarities and differences as regards the areas where administrative burdens are likely to be most significant. To be sure, priority areas for simplification in each member state may differ depending on the industry structure and peculiar legal, social and economic features. Member states are not homogeneous in all these dimensions: as a consequence, the mix of administrative costs imposed on businesses and the cost reduction potential for each state cannot be the same. This caveat applies both to the comparison of the results obtained by the four member states analysed, and even more importantly to the extension of the identified list to all other member states. In Section 4 below we will briefly address the issue of how to deal with the extrapolation of data from a limited number of member states to all the EU25. Secondly, it has to be emphasized that the evaluation of the priority areas depends on what national governments decided to include in their measurements. In the case of EU legislation, which falls outside the sphere of influence of national administrations, the corresponding IOs/DRs might not have been completely measured by member states. This is understandable, as so far the implementation of the SCM tool in these countries was undertaken exclusively for the purpose of setting national reduction targets, which necessarily do not include reduction of costs generated by non-national legislation. Thirdly, even after selecting a list of most burdensome areas, it remains difficult to draw any conclusions as regards the reduction potential in each of these areas, since: a) not all administrative costs are unnecessary; and b) less costly alternatives are not always available. For this reason, the identification of a list of priority areas must be followed by consultation with member states, in order to identify suitable proposals for simplification of EU legislation within each of the areas. In this respect, we recommend that simplification priorities are identified in close collaboration with national administrations. Finally, from a methodological viewpoint, substantial problems emerged in identifying most burdensome areas, as each member state used a different breakdown of policy areas when classifying burdensome IOs/DRs. A typical feature of existing national implementations of the SCM is that IOs/DRs are attributed to ministries/departments, and the distribution of administrative costs amongst different areas can reflect the different division of competences between ministries/departments at national level. In other words, different pieces of EU-legislation were compiled according to national specific definitions. – 114 – In summary, we adopted a ‘bottom-up’ approach in identifying priority areas for simplification. This approach has both advantages and drawbacks. On the one hand, we were able to identify priority areas based on actual results obtained by member states through their baseline measurements. On the other hand, each country provided its most burdensome areas according to the specific national definition of policy areas/domains, which inevitably differ. Accordingly, the figures we received are not fully comparable. In any event, when analysing available data from national measurements, we identified some common evidence as regards most burdensome areas. The set of priority areas was then generated through listing comparable domains/areas as countries defined them and summing them up to priority areas. The list of priority areas was then transmitted to the participating Better Regulation units for validation, to verify the comparability of the compiled areas and to clarify whether burdensome areas were left out. When selecting priority areas for simplification, not only the number of the burden should be considered, but also the ‘irritation factor’ on businesses – especially SMEs – has to be taken into account. For example, the Dutch IPAL proposes to consider also environmental legislation and licenses amongst priority areas. Table 30 below shows the Dutch top-10 list of areas – which SMEs understand as most irritating – and their origin. Table 30 - Irritating areas for SMEs and their predominant origin according to IPAL 1. Statistics: mostly European law 2. Safety at the workplace: mostly European law 3. Tax declaration: national legislation 4. Personnel bookkeeping: mixed 5. Environmental legislation: mostly European law 6. Licenses: mixed 7. Bookkeeping for driven Km by car: national legislation 8. VAT: mostly European law 9. Repro law: national legislation 10. Annual accounting: mostly European law Based on all the abovementioned considerations and caveats, the key priority areas we identified are the following: • Annual accounts and company law; • Health Protection (including animal health and zootechnics) • Working Environment and employment relations; • Fiscal Law and VAT; – 115 – • Statistics127. • Agriculture and agricultural subsidies128; • Food labelling; and • Transport. Importantly, a further proposal was made by the Danish Division for Better Business Regulation: Denmark proposes to include also the financial services area. Experiences according to this area were made through a benchmark too. Several countries seem to exhibit significant burdens in this area through EU-legislation. In the UK, as reported in Section 2 above, financial services were subjected to a separate measurement, which yielded a (conservative) estimate of £440 million. Detailed data on the percentage of administrative burdens accounted for by EU legislation in this areas are not available for the UK. Table 31 below provides examples of burdensome pieces of EU legislation in each of the selected priority areas and illustrates the available information as regards the relative percentage of total administrative costs accounted for by each area in Category A and B in the four member states analysed. 127 The area Statistics does not show the large burden as for example Annual Accounts, but it reportedly causes significant irritation, which can be seen as a proxy for the existence of unnecessary burdens: hence, we deem it appropriate to include this area in the list of priority areas. 128 The Netherlands, the United Kingdom, Sweden and Denmark are working on a benchmark project on the agricultural sector. The draft outcomes so far show significant burdens in this domain of EU-legislation. More details are not available at this stage, but it seems advisable to include the agricultural sector in the list of priority areas for EU- wide measurements. – 116 – Table 31 – List of priority areas – 117 – – 118 – 3.3 Most burdensome EU directives The results of the four measurements undertaken in the Netherlands, Denmark, the UK and the Czech Republic allow, with some adjustment and additional work, the identification of pieces of EU legislation that are particularly burdensome for private enterprises. Again, the extent to which the results of the measurement can be considered comparable and homogeneous is limited by the differences in the scope and methodology used by the four governments in implementing the SCM. Table 32 below contains a list of most burdensome EU directives that was compiled according to data provided by the central coordinating units in the Netherlands, Denmark and the Czech Republic.129 The results were validated with the help of UK data at a later stage, as reported in the right end column. Estimated figures for some areas are reported as ‘not available’ – in particular as regards Denmark and The Netherlands – only due to time constraints. As emerges from the table, some directives have been estimated as considerably burdensome in all member states. For example: • Council Directive 78/660/EEC of 25 July 1978 on the annual accounts of certain types of companies (the “Fourth Directive”) generates significant burdens both in Denmark (€409 million) and in the Netherlands (€1,497 million); • Council Directive 77/388/EEC of 17 May 1977 on the harmonization of the laws of the Member States relating to turnover taxes - Common system of value added tax: uniform basis of assessment (the “Sixth Directive”) causes ‘only’ €43 million of administrative costs in Denmark, but as high as €1,427 million in the Netherlands. In the UK Tax Measurement, EU Directives on VAT were found to account for more than £900 million.130 • Directive 2000/13/EC on the approximation of the laws of the Member States relating to the labelling, presentation and advertising of foodstuffs generates €93 million of administrative costs in Denmark, €254 million in the Czech Republic and €316 million in the Netherlands. This is reflected by available data in the UK. • Directive 89/391 of 12 June 1989 on the introduction of measures to encourage improvements in the health and safety of workers at work generated over €145 million of administrative costs in Denmark. • Directive 92/57 of 24 June 1992 on the implementation of minimum safety and health requirements of temporary or mobile construction sites generated over €54 million of administrative costs in Denmark. 129The UK BRE has not provided data on most burdensome directives, but allowed access to the database at a later stage. As the final results will be published by the end of 2006, we do not report them in this version of the Final Report. 130 See KPMG, Administrative Burdens, HMRC Measurement Project, 20 March 2006, pp. 24-25. – 119 – Table 32 – most burdensome EU directives, cross-country – 120 – 4 SUGGESTIONS FOR THE COMMISSION’S “ACTION PROGRAMME” In the previous sections we have analysed the similarities and differences between the SCM tools adopted by the four member states that have completed a baseline measurement to date. We also highlighted the current differences between these national models and the EU SCM, the model currently adopted by the European Commission for the measurement of administrative costs stemming from a number of new proposed legislative initiatives, which will also be applied in the upcoming Action Programme on administrative burdens to be launched by the Commission in 2007. The timetable for the upcoming Action Programme is shown below, in Figure 10. After completion of this Pilot Project, The European Commission will issue a Communication on the reduction of administrative costs, which will be followed by a consultation process and by the launch of the Action Programme in February 2007. Later, in April 2007 the Commission plans to launch the (partial or total) measurement exercise. The end of the measurement exercise is scheduled by the end of 2008, followed by the launch of cost reduction strategies. Figure 10 – Timetable of the upcoming Action Programme 2006 2007 2008 Communication Action Launch of Targets – action for on AC Cost Programme on measurement and decision and Reduction Reducing AC reduction exercise implementation November February March April December Measurement • Discussion on • Methodological • Council the results of issues conclusions on the pilot project Action • Organisational Programme and on targets issues • Launch of • Targets on Low • Identification of Hanging Fruits consultation on policy areas Low-Hanging • Launch of Fruits and • Identification of measurement priority policy Low-Hanging and reduction areas Fruits initiative As we already mentioned in Section 1, a number of issues still remain to be tackled for the EU SCM to be effectively and successfully implemented in a full-scale – or at least – 121 – multi-sectoral – measurement in combination with national models. These have been summarised as follows. • Database interoperability: to ensure the interoperability of national databases on administrative burden and access for the Commission. • Country distribution: identification of weighting systems for assessing EU-wide costs on the basis of a limited number of national data (e.g. country distribution). • Accuracy: identification of the average margin of error of administrative cost assessments. • Standard ratios: identification of standard ratios for overheads, training costs and learning curves and for costs corresponding to normal business operation, among other things. • Thresholds: identification of specific threshold(s) below which quantification is not necessary (minimum thresholds for the application of the model). • Extension to citizens: possible adjustments of the model when assessing administrative costs put on citizens. • Guidance on borderline cases: possible difficulties to distinguish information obligations from the other regulatory costs and how to overcome them. • Standardisation of IOs and target groups: looking at possible shortcomings of the typologies of IOs and required actions used in the EU operational manual; examining the need for a typology of target groups. • Exchange of data: Organising optimal exchange of data between the Member States (including their regional authorities) and the Commission. • Target-setting: the issue of target-setting still needs to be addressed and agreed upon. Whether an overall target can be set at pan-European level for the reduction of administrative burdens imposed by legislation within a given timeframe is one of the issues that will have to be explored in the near future. Besides these issues, the implementation of a pan-European measurement exercise will also exhibit a number of operational problems, which have to be solved before such an ambitious endeavour is undertaken. These are mostly related to the possibility of communication between the EU and national administrations and databases, to the need to adapt national models, to the division of tasks over the different phases of the SCM tool, to target-setting issues and to institutional issues such as whether and how to ensure that all EU institutions commit to a common cost reduction target. Therefore, virtual networks like SINAPSE might be useful, although they cannot replace face-to- face meetings. One of the main objectives of this Pilot Project is to provide first suggestions and guidance on how to proceed in the upcoming Action Programme. We proceed as follows. Section 4.1 discusses methodological issues, by identifying different scenarios for the implementation of the EU SCM in the Action Programme. This includes an – 122 – identification of possible solutions for the classification of origin, possible arrangements for the communication and interoperability between EU and national databases, methodological problems in data collection, sampling and extrapolation, and the application of the principle of proportionality. Section 4.2. deals with organisational issues, such as a description of ‘who does what’ in a multi-level environment where the Commission and member states are jointly involved in the measurement. In this section, we provide a step-by-step analysis of the measurement process, by identifying potential problems that may emerge in the different phases of the SCM tool when implemented in the Action Programme. This section implicitly deals with the application of the principle of subsidiarity in the implementation of the EU SCM. Section 4.3 contains more operational suggestions for the upcoming sectoral measurement, such as the scope and purpose of the upcoming measurement, the priority areas that could be selected, alternative cost-reduction strategies, budget allocation problems, and a list of activities that should be carried out by consultants when preparing the measurement. We also deal with the delicate issue of setting overall and sectoral targets for the reduction of administrative burdens in the EU25, and provide some hints on how to involve all EU institutions in the commitment to reach the overall/sectoral targets that will be set in the Action Programme. Finally, Section 4.4 summarises our main findings. 4.1 Methodological issues In Section 3, we highlighted the main similarities and differences between the national variants of the SCM in the four countries that completed the baseline measurement to date. Our main conclusion was that the adoption of a common theoretical framework will certainly prove essential for the development of a pan-European cost reduction strategy. This, in turn, implies that all countries that have not launched a measurement exercise to date should adopt a model that shares most of the main features with those analysed in this Pilot Project. In any event, as we have shown, 7 of the EU25 have undertaken or plan to undertake a full baseline measurement, and other 10 member states are carrying out at least a pilot measurement exercise in some policy areas. In this respect, the launch of a pan-European Action Programme will imply that the Commission takes action to ensure that remaining member states also adopt a cost reduction strategy in the areas that will be selected. Otherwise, the goal of setting targets at both the EU and national level would not prove attainable within the Action Programme. In this section, we discuss some of the pending methodological issues as identified by the European Commission. To avoid redundancies, our discussion draws on the description of the current EU SCM as provided above, at Section 1.7. The issues addressed by this section include the classification of origin, achieving database interoperability, how to collect data and identify samples of businesses, and how to extrapolate collected data at pan-European level. – 123 – 4.1.1 Classification of origin An important issue in the implementation of the SCM is the classification of IOs/DRs according to their origin. As shown in Section 2, all countries analysed – and also other countries that have started to implement the SCM tool – adopt the ABC(D) classification.131 Of these, two countries – the UK and Denmark – reached a more detailed classification, by distinguishing categories A and B in three sub-categories, i.e. “EU Directives”, “EU-Regulation” and “Other International”. On the other hand, the EU SCM proposes a different taxonomy, which distinguishes between “International”, “EU”, “National” and “Regional”. A key feature of the upcoming Action Programme will be the setting of targets for cost reduction at both EU and national level. In this respect, achieving a clear distinction between IOs/DRs generated by EU legislation and national legislation is of utmost importance. As the classification of origin is meant to identify IOs/DRs that fall under the sphere of influence of a given legislator (EU, other international, national, regional), a mistaken classification would shift the responsibility for cost reduction from one legislator to another, thus creating potential inconsistencies and the impossibility to tackle effectively the problem of cost reduction. The main problems that have emerged in this Pilot Project as regards the classification of origin can be summarised as follows: • The ‘simple’ ABC classification is not suitable to clearly identify IOs/DRs generated by EU legislation, as the latter are merged with IOs/DRs imposed by other international legislation. These countries have declared that non-EU international legislation is mostly to be considered as a residual category, accounting for less than 10% of total administrative costs. In the UK, however, the ‘A-other-international’ and ‘B-other- international’ categories seem to account for a greater share of total administrative costs in Categories A and B with respect to EU Regulations. In Denmark, to the contrary, the total figure of how much the category ‘EU other‘ constitutes of A and B regulation is 2,48%. In other words, a clear-cut distinction between EU and other international sources of legislation must be achieved if targets are to be efficiently set in the Action Programme. As a result, the European Commission should ensure that countries that still have to undertake the measurement exercise and countries – e.g. the Netherlands – that are approaching a second baseline measurement adopt a more detailed classification, at least similar to that adopted in the UK and Denmark. • In some cases, the distinction between Category A, B and C is difficult to establish. Examples from the benchmarking exercises carried out by the International SCM network and evidence from national databases (e.g. the UK) reveal that in many instances the attribution of a given IO/DR can prove difficult, leading to a significant degree of discretion in the classification of origin. In particular, 131The four countries analysed do not have regional legislation. Countries with regional structure that have started implementing the SCM tool, such as Italy, have introduced a ‘Category D’ which includes IOs/DRs originated by regional legislation. – 124 – distinguishing between IOs/DRs that originate from the text of an EU Directive and those that arise from the national implementation of the Directive can prove quite difficult. The same applies to cases in which the reference to the original EU piece of legislation is missing in the national implementation measure. For such reasons, as will be clarified in more detail in section 4.2 below, the Action Programme should attribute to the European Commission (o its consultants) the task of identifying all IOs/DRs generated by the acquis communautaire, thus creating the ‘EU’ category of administrative costs. • In national models analysed, the issue of identifying Category A IOs/DRs was certainly less important than in the upcoming Action Programme. This is understandable, as Category A administrative burdens fall entirely outside the sphere of influence of national legislators, and thus are not included in national cost reduction targets. In this respect, the UK database proves particularly useful, as in practice most Category A burdens were actually measured and included in the database.132 However, the availability of a benchmark database will prove useful, but not decisive for the EU mapping of the acquis. As observed during the Pilot Project, in some cases the reference to the original piece of EU legislation is missing in the UK database, and the distinction between A-EU-Regulation and A-EU-Directives is not clearly established. This does not affect at all the accuracy of the UK national measurement, but may prove important for a clear-cut mapping of the acquis and the identification of actions needed to reduce administrative costs generated by EU legislation in the short- or medium term. • The ABC(D) classification of origin does not always divide the responsibility between EU and national legislators clearly. This occurs mostly whenever EU pieces of legislation require some sort of implementation at national level, without specifying which measures should be adopted. This is the real ‘grey area’ between EU and national legislation. For example, imagine that an EU Directive introduces a control system on businesses that will the require some inspection and reporting activities, but does not specify the frequency of inspection, nor that of reporting. In this case, national legislators cannot avoid introducing further measures to specify how the new rules will be made operational at national level. However, implementation measures can be set at a minimum level - e.g., quarterly inspections and yearly reporting – or at a more burdensome level – e.g., monthly inspections and quarterly reporting. As some measure must be adopted by the national legislator, the former case cannot be considered to belong to the sphere of influence of national legislator, whereas the latter is the result of a national choice. However, both sets of measures would be classified as ‘B’. The abovementioned problems lead us to believe that the classification of origin should be carefully addressed by the European Commission during the preparatory phase of 132More precisely, the UK database may serve as a useful benchmark for the European Commission in mapping all the IOs/DRs originated by the acquis for private and semi-private businesses, but not for citizens – as the UK did not include administrative costs for citizens in its baseline measurement. – 125 – the Action Programme. Figure 11 below describes the process of building what is sometimes termed the ‘zero’ Category, or the ‘EU’ Category in the EU SCM. Figure 11 - Building the 'EU' Category Minimum Minimum implementation implementation A B C D Directive Directive National Regional EU National National Regional Category ‘EU’ Regulation Regulation Category Other-Int’l Other-Int’l International ‘International’ Following these considerations, some alternative options can be identified for achieving a classification of origin which proves functional to the peculiar needs of the upcoming Acton Programme. • Option 1: keeping the (detailed) ABC(D) classification as it is This option would prove less intrusive for member states that have already started measuring administrative burdens or have completed the baseline measurement. However, member states that have applied a simple ABC classification would have to re-start their measurement of category A burdens. • Option 2: ABC(D) classification, with Category A defined by the Commission Such option would imply that the Commission provides its own list of IOs/DRs to be included in the A Category, so that the measurement is standardised to the extent possible. A similar option, however, would not represent a suitable solution for Category B, which contains IOs/DRs that fall partly under the sphere of influence of the Commission. In the pan-European measurement exercise, constant checks for the consistency of IOs/DRs selected by member states would be needed, especially in Category B, to avoid double counting and cases of mistaken classification of administrative burdens. Operationally, this option would imply that: a) the Commission defines all Category A(EU) IOs/DRs; b) member states define residual category A(International) IOs/DRs; c) member states measure categories A, B and C. This division of tasks might create inconsistencies in the organisation of the measurement at pan-European level. In particular, this option can create the problem of identifying IOs/DRs that correspond to the minimum implementation of an EU piece of legislation, and those that go beyond the minimum level of implementation. – 126 – • Option 3: adopting the EU SCM’s four categories, with definition of Category ‘EU’ by member states In this case, the European Commission would rely on the results of national measurements to build the category ‘zero’ or ‘EU’. This option, however, would exhibit at once:: a) a lack of standardisation, since category zero would be built through the input of member states, which may classify IOs/DRs differently according to their peculiar implementation of EU legislation and to the decisions on classification taken by ministries and consultants; and b) costs for member states that have adopted the ABC(D) classification, as they would have to update their original databases by changing the current classification into the EU one, and to formulate assumptions on the minimum efficient level of implementation of EU directives and regulations, in order to divide responsibilities between the EU and national sphere of influence. • Option 4: adopting the EU SCM’s four categories, with definition of the ‘EU’ category by the Commission Under this scenario, the identification of Category ‘EU’ would be undertaken at EU level, with the help of consultants. The main activities to be carried out at EU level would be: a) the analysis the acquis and the identification of IOs/DRs directly generated by prescription of current EU legislation; and b) assumptions on the ‘minimum efficient level’ of implementation of those pieces of EU legislation that do not specify them directly. Such assumptions should then be validated by member states before a final list of IOs/DRs under the ‘EU’ category is reached. • Option 5: adopting the EU SCM’s four categories, with definition and measurement of the ‘EU’ category by the Commission Under this scenario, the Commission would have to undertake the measurement of the IOs/DRs included in Category ‘EU’. This option would produce the following consequences: a) high standardisation, as the measurement would be completed at central level; b) very high costs for the European Commission, as the measurement exercise may prove prohibitively resource-intensive, and was estimated in the magnitude of 100 million; c) redundancies in the measurement with respect to national measurements, possibly resulting in duplication of interviews, focus groups, etc. For this reason, this options seems less desirable than option 4. Table 33 below compares the five options identified. – 127 – Table 33 – Options available for the classification of origin Option Costs for EU Costs for MS Standardisation 1 – Status quo low low none 2 – Commission defines ‘A’ medium low low 3 – Four categories, MS define ‘EU’ low high none 4 - Four categories, EC defines ‘EU’ medium medium high 5 - Four categories, EC measures ‘EU’ high low high In summary, option 4 seems to allow for the maximum standardisation, which is highly needed in light of the upcoming Action Programme, and divides the cost of the measurement more equally between the EU and national level. This option would imply that significant costs are borne the EU level, as the identification of category ‘EU’ would require both a screening of the acquis and assumptions on the minimum level of implementation of EU pieces of legislation. In achieving this results, as already recalled, Commission DGs (or consultants) might rely on databases such as those developed in the UK after the baseline measurement, although this source does not always allow to identify regulatory origin in an uncontroversial way in Categories A and B. On the other hand, member states would then be able to: • easily define Category B as a residual category starting from the EU identification of the ‘EU’ category; • easily identify instances of gold-plating; and • maintain their classification of origin for part of category B, category C and – where existing – D.133 In this respect, it must be recalled that member states so far have not committed to undertake the measurement at national level, but only to provide the Commission with all available data and information needed to perform the measurement at central level. In other words, there is still no basis for selecting option 4 in official documents approved at Council level. In case such an agreement proves impossible to reach, a mix between options 4 and 5 would seem advisable, whereas the Commission obtains or reuses data from member states that have already measured given IOs and performs the measurement (or extrapolates) directly in countries where the measurement has not yet been undertaken. 133Member states would then be free to decide whether to keep the distinction between B and C or to merge the two categories into the ‘National’ one, as foreseen by the EU SCM. – 128 – 4.1.2 Database interoperability A related issue that might emerge during the EU-wide measurement exercise is how to ensure that the EU and national databases are made interoperable. This issue would be adequately tackled if the Commission decides to carry out the task of defining the EU category (option 4 above). In this case, the Commission would directly provide member states with a list of IOs/DRs that should be considered under its sphere of influence. However, the problem of interoperability would emerge at a later stage, when the list of IOs/DRs belonging to the ‘EU’ category is actually measured. At this stage, the responsibility for the measurement exercise should be carefully coordinated and highly standardised. As the Action Programme will imply the setting of targets both for the EU and member states, measurement should be undertaken at both levels. This creates a further trade-off: responsibility for the measurement – including interviews, business panels, focus groups etc. – should be allocated either at EU or at national level. If the measurement is undertaken at EU level (option 5 above), the problem of interoperability would be partly solved, as the EU would carry out its own measurement in a fully independent and autonomous way. This option, however, would create significant problems, such as: • Redundancies in the measurement exercise – e.g. duplication of interviews, focus groups, etc. • Possible inconsistencies in the identification of the ‘normally efficient business’, let aside the measurement for citizens; • Possible inconsistencies between the EU measurement and national measurements – e.g. double counting of IOs/DRs or, at the other extreme, failure to measure borderline IOs/DRs which are considered ‘national’ by the Commission, and ‘EU’ by member states. For such reason, we confirm our suggestion that the measurement of the ‘EU’ category would have to be carried out by member states and coordinated by the Commission or its consultants in order to avoid inconsistencies in the procedure. This option has the further advantage of saving the efforts of those member states that have completed the baseline measurement so far, provided that they have identified IOs/DRs originated by EU legislation in a consistent way. Those member states that are planning to launch a full or sectoral measurement in the near future should then adapt their own models according to the peculiarities of the model adopted in the upcoming Action Programme. A further related issue is that of the so-called ‘table of concordance’ between EU and national legislation. The functioning and operation of the forthcoming EU database would obviously be smoother, where a clear table of concordance between EU and – 129 – national legislation could be available.134 In order to be effectively implemented in the forthcoming Action Programme, such tables would need to clearly indicate: • What pieces of national legislation implement what piece of EU legislation; • Which parts of national laws/regulations – articles, sections, etc. – are the direct effect of transposition of EU Regulations/Directives, and of which EU Regulations/Directives. • A full correspondence between IOs/DRs originated by EU legislation and their location in national legislation. There are two main options available to reach this result, namely a ‘bottom-up’ and a ‘top-down’ approach. • Bottom-up approach: Member States provide the table of concordance, after a careful screening of their own legislation. This basically requires that all member states undertake such exercise, which is normally part of the implementation of the SCM tool, with some further complication as regards the classification of origin.135 • Top-down approach: the Commission provides a list of IOs/DRs generated by the acquis, whereas member states trace them, flag them and classify them as ‘EU’ in the national implementation of the SCM tool. The table of concordance is spontaneously created at the end of the exercise. Also in light of our conclusions in the previous section on the classification of origin, we believe that a top-down approach would ensure a greater level of consistency and a clearer division of responsibilities between the EU and member states. Finally, an issue that is often raised is whether it would be desirable to develop a pan- European database of administrative burdens, encompassing figures on administrative burdens from all the EU25. Representatives of national coordinating units – in particular the Danish Division for Better Business Regulation – were rather sceptical on this perspective. An application of the proportionality principle in this case would rather suggest that the Commission initially focuses on a limited number of countries when collecting data to reach estimates of total administrative burdens in given sectors. As will be explained below, we suggest that the Commission works on the pan-European database at a later stage, when (and if) all member states will have launched full or at least sectoral measurements based on the SCM tool. 134Accordingly, for the purposes of this report, concordance tables should therefore not be confused with the ‘light table of concordance’ referred to in the Commission Staff Working Document, Annex to the Communication on Better Regulation for Growth and Jobs in the European Union, ‘Minimising Administrative Costs Imposed by Legislation’, SEC(2005)175, p. 13. In order to fully implement the SCM tool in a multi-level measurement exercise, the concordance table would have to be way more detailed that that included in the reporting sheet of the EU SCM, as it would have to be broken down at IO level. 135In most cases, national databases analysed do not always provide a reference to the original piece of EU legislation, both for Category A and B. – 130 – 4.1.3 Sampling, data collection, extrapolation The SCM tool was initially conceived and implemented at national level. When implemented at EU level, some of its major phases require a careful adaptation. In particular, the identification and segmentation of the population, the identification of the normally efficient business (or ‘administration’, or ‘citizen’), the choice of a representative sample, the use of data collection methods and the extrapolation at EU level become particularly delicate. In this section, for sake of simplicity, let us assume that option 4 in Section 4.1.3 and the ‘top-down approach’ in Section 4.1.4 have been selected. This means that the Commission identifies Category ‘EU’ and member states carry out the measurement at national level with the help of their consultants under the supervision and coordination of the European Commission. In this case, after identifying the IOs/DRs to be measured, member states should look at the population affected by those IOs/DRs by identifying the type of businesses affected, possible segmentation arrangements and the normally efficient firms. This can be achieved with the help of consultants which follow the methodological instructions contained in the international SCM manual and incorporated in the EU SCM operational manual, and also on the basis of national baseline measurements already completed by member states. One important feature of this procedural arrangement is that ‘normally efficient entities’ may be defined differently by member states. As a result, there would be no need to define a single ‘normally efficient entity’ for the whole EU25 – an exercise that would otherwise prove particularly delicate. A possible shortcoming is that the overall calculation of administrative costs and their distribution by country may be affected by differences in the assumptions adopted at nationals level as to what constitutes a normally efficient entity. For this reasons, we suggest that the European Commission (or its consultants) strongly coordinates and supervise this procedure, or alternatively provides a preliminary estimate of time and other costs needed to comply with IOs/DRs in the EU category, which member states are then called to validate by presenting the results of their interviews, focus groups, panels or expert assessments. Such ‘reversal of the burden of proof’ might help achieving a smoother coordination between the EU and national level when undertaking the measurement exercise. In addition, we consider it quite unlikely that the Commission ends up relying on all the EU25 when collecting estimates of category ‘EU’. If this were the case, extrapolation to the EU level would result automatically by summing up the results of extrapolations to the national level. However, this requires that member states carry out their measurement simultaneously, at least in the priority areas that will be selected.136 A less burdensome and ambitious option would imply that the Commission relies on a limited number of member states, and then extrapolates the results at EU- wide level. For example, the Dutch CPB has provided an estimate of total 136 Member states that have already measured – 131 – administrative burdens in the EU by extrapolating the Dutch figure and using cost differences to reach an identical standard firm in each Member State. Table 34 – Example of country distribution (CPB, 2006) Possible parameters that could be used for extrapolating data to the EU level using evidence available from some member states differ depending on the type of policy area and type of IOs/DRs under analysis. A recent example of extrapolation exercise carried out by the European Commission services can be found in the “Impact Assessment of the Thematic Strategy on the Sustainable Use of Pesticides”.137 In the recent “Impact Assessment of the Thematic Strategy on Soil Protection”, to the contrary, extrapolation was not considered possible in a number of instances due to lack or reliable data.138 We suggest that, in case data from a limited sub-set of EU member states is available, the Commission proceeds with extrapolation in much the same way, and then asks member states (or, as will be proposed below, to the International SCM Network) to validate the findings. Finally, in case the measurement exercise has to be undertaken at EU level (option 5 above), data collection methods commonly used at national level – e.g. face-to-face interviews – may prove too costly. The UK experience suggests that business panels can prove particularly cost-effective in providing reliable data estimates. At EU level, use of the European Business Test Panel might prove particularly useful. The EBTP is composed of around 3.600 companies of different sizes and sectors located in all EU Member States, which reply to online questionnaires sent to them by the European Commission on a regular basis. In other instances, industry and trade associations may be consulted directly through telephone or even face-to-face interviews. Another important lesson learnt from the UK experience is that a careful implementation of the principle of proportionate analysis would be needed to carry out the measurement at EU level. As explained in the previous sections, the results of national measurements completed to date highlight that often a very small percentage of the IOs/DRs identified accounts for a very large percentage of the resulting administrative costs. For example, the top 20 IOs/DRs originated by EU Regulations in the UK represent 95.46% of total administrative costs originated by EU Regulations, and 94.33% of net costs. This suggests that for a large portion of IOs/DRs, direct expert assessment may be a practical and efficient solution. As reported above in Section 2.3, direct assessment was used in the UK for 55.7% ot IOs/DRs, accounting for 7.4% of total administrative costs. In summary, it is fair to conclude that problems of data collection and sampling may become almost insurmountable under option 5 – i.e., if the Commission doe not only 137 COM(2006)373 Final, 12 July 2006. 138 COM(2006)231 Final, 22 September 2006. – 132 – have to define category ‘EU’, but also has to measure it. Alternatively, if option 4 is chosen, these methodological problems would mostly be related to careful selection of representative member states and to the extrapolation of results. On these two issues, we suggest that: • Constant collaboration with the International SCM Network is ensured when selecting representative member states; and • The Commission continues its research on extrapolation criteria, as are currently being used in ex ante impact assessment. These two suggestions, in our opinion, would respect the principle of proportionality and subsidiarity, leaving member states with a task they are supposed to perform better – identifying the costs, time, population, segments and ‘normally efficient entity’ according to national peculiarities – and awarding the Commission the role of defining the ‘rules of the game’ and extrapolating data to the EU level at the end of the measurement. 4.1.4 Level of detail Another important issue to be tackled before proceeding with the Action Programme is the level of accuracy and detail that the EU measurement should reach. In the case studies illustrated in Section 2, we highlighted that countries that have broken down IOs into data requirements (e.g. Denmark) have reached a higher level of accuracy. However, observation of the UK database has also revealed that the correspondence between IOs and DRs was almost one-to-one in the measurement. Also the Danish Division for Better Business Regulation confirmed that the higher level of detail reached in Denmark would be hardly replicable in larger countries. For such reasons, we suggest that the Commission focuses on IOs and only subdivides IOs into DRs when there is sufficient evidence that a single IO requires a more detailed breakdown. 4.2 Organisational/Institutional issues In this section we address the most relevant institutional and organisational arrangements that appear necessary in light of the forthcoming Action Programme. The most important issues are, in our opinion, the identification of the scope and purpose of the measurement exercise, the division of tasks between the EU and national level, and the involvement of other EU institutions in the operation of the EU SCM and in the commitment to reach the targets that will be set within the Action Programme. 4.2.1 Scope and purpose As we already mentioned in Sections 1 and 3, the EU SCM was conceived with a very ambitious scope, i.e. measuring administrative burdens for private and semi-private enterprises, but also for citizens and administrations. This feature makes the scope of the EU SCM broader than that of available national models: as a matter of fact, only in – 133 – the Netherlands the extension to citizens was introduced, and in none of the four countries the costs borne by public administrations was included in the measurement. Moreover, in the Netherlands the (simple)ABC classification was used, and in the Dutch database tracing back the original piece of EU legislation is not always possible. This, in turn, also means that the Commission will not be able to rely on any existing model when identifying the IOs/DRs for citizens and public administrations. In this respect, we suggest that the Action Programme initially focuses on private and semi-private enterprises, whereas administrative costs for citizens and public administrations may be considered at a later stage and within the ex ante impact assessment of new pieces of legislation. The Dutch experience also suggests that the measurement of administrative burdens for citizens should be kept separate from that regarding businesses. 4.2.2 ‘Who should do what’ in the Action Programme In the previous sections we have formulated some initial suggestions on how to approach the upcoming measurement exercise at EU level. In this section, we provide a step-by-step analysis of the activities that should be performed by the European Commission, by Member States and by consultants. Table 35 below shows a possible allocation of tasks, by taking the Danish model as a reference (the columns on the right) and by assuming that the EU and national measurement take place simultaneously. Of course, the table must be considered as a mere example. – 134 – Table 35 – Example: ‘who should do what’ in the Action Programme As shown in the table: • in ‘phase 0’ Commission DGs and their consultant(s) should identify the business related regulation to be included in the analysis. This might imply the selection of priority areas and the identification of DGs that should be involved in the Action Programme. Such process is very similar to what occurs at member state level, where some of the ministries are often not included in the measurement as they do not produce a significant amount of business-related regulation. • In step 1a, Commission DGs and their consultant(s) should map the acquis in order to define IOs/DRs to be included in category ‘EU’, including those representing the minimum efficient implementation of the pieces of EU legislation analysed. • In step 1b, central units, ministries and consultants would be identifying the IOs/DRs generated by other international and national legislation. This activity should ideally be undertaken after the Commission has identified category ‘EU’, so that member states can define their own category ‘B’ as a residual category under the sphere of influence of national government. – 135 – • In step 1c, the Commission’s definition of category ‘EU’ should be validated by member states, possibly by ministries. Involvement of central coordinating units would also be essential, possibly through the International SCM Network. • In step 2, central units and consultants should deal with demarcation both at the EU and national level (for non-EU categories). • In subsequent steps, (a limited number of) member states should measure both category ‘EU’ and all other categories, exactly as they would have done absent the Action Programme.139 However, the identification of cost parameters for ‘EU’ IOs/DRs and the preparation of the interview guide should be be carried out in collaboration with the Commission’s consultant. • In step 9, ministries (in our representative model, i.e. Denmark) and Commission DGs (only for the ‘EU’ category) should deal with the expert review of previous steps. • During phase 2, EU consultants should participate to step 11 (Completion and standardisation of time/resource for each segment by activity), whereas Commission DGs should deal with the expert review of previous steps (only for the ‘EU’ category). • In phase 3, national consultants extrapolate data to national level, whereas EU consultants deal with aggregation and extrapolation of data t the EU level. Commission DGs and their consultants then report results in the EU database, whereas ministries and national consultants report their results in national databases. 4.2.3 Institutional communication issues Another issue that was included in the areas for optimisation of the EU SCM model is the need to ensure constant collaboration between the Commission and competent administrations in member states at national and regional level. The implementation of the Action Programme will require that the interface between activities performed at Commission level and those carried out by member states are efficiently coordinated. In this respect, the following suggestions can be formulated, under the assumption that the options highlighted above are selected by the Commission: • During the sectoral measurement that will be launched, the Commission will need to secure that representatives of central coordinating units at national level are available to validate the results of phase 1b, thus reaching an agreement on the list of IOs/DRs that should be considered as ‘EU’. • When performing a sectoral measurement, the competent Commission DG and competent ministries at national level should reach an agreement on the population, the 139 The only difference is that they would have to measure all the ‘EU’ and ‘Other international’ IOs/DRs identified. – 136 – sampling, data collection methods and cost parameters. This agreement should lead to the development of standard ratios overtime. • In addition, in this same phase the Commission would need to ensure that national governments can produce reliable data in a timely and precise manner, for example consulting offices of national statistics and other official sources of data. • After the measurement, national ministries should validate the results of the extrapolation to EU level carried out by the DGs with the help of consultants. To the contrary, if the Commission ultimately has to measure administrative costs directly, but using data provided by (a limited number of) member states, the degree of coordination and communication would prove more challenging. Moreover, effective communication between the Commission and member states could be achieved – at least for those phases that require communication between central coordinating units and the Commission – by achieving constant collaboration between the Commission and the International SCM Network, which gathers representatives of central coordinating units of those governments that have already officially adopted the SCM tool for cost reduction strategies. Such communication has proven to be quite effective already during this Pilot Project, which could strongly rely on data provided by the four member states analysed. In the past few months, a growing number of EU member states joined the SCM Network. “Institutionalising” such collaboration would also imply that, as the participation of EU member states to the SCM Network becomes more numerous, the Commission would consequently be able to rely on a growing number of national counterparts when performing the various phases of the EU SCM. Finally, use of web-based tools such as the newborn SINAPSE would prove essential to speed up consultation processes between the Commission and representatives of central coordinating units at national level. This tool can also enable constant confrontation on the development of standard ratios and parameters, e.g. overheads, time estimates, etc. 4.2.4 Involving all EU institutions As already recalled at the beginning of this section, the forthcoming EU Action Programme will imply the formulation of cost reduction strategies and the setting of overall/sectoral targets for reducing administrative costs within a given timeframe. We will discuss this issue in more detail below, at Section 4.3.2. Whatever option is chosen for target-setting, no target can be achieved at EU level if only the Commission commits to successfully implement the cost reduction strategy. As a matter of fact, implementing a cost-reduction strategy will certainly imply a rule-making effort at EU level, aimed at amending and/or repealing particularly burdensome pieces of legislation. Thus, targets can be achieved only if the European Parliament and the Council also commit to cooperate in achieving the target when deciding on proposed new pieces of legislation within the co-decision procedure. – 137 – The most straightforward way to achieve this result would be to amend the current Inter-Institutional Agreement on Better Lawmaking to include targets set within the Action Programme. The Inter-Institutional Agreement implies, i.a., that all EU institutions share the same impact assessment methodology: as the EU SCM has been integrated as an annex to the Commission’s Impact Assessment Guidelines, it is fair to state that the EU SCM is already part of the Inter-Institutional Agreement. However, targets are not. In this respect, we suggest that the Commission, the Parliament and the Council reach a new Inter-Institutional Agreement, in which all institutions commit: • To formulate their decisions – including initiatives by the Commission and amendments by the Parliament and the Council – by awarding priority to the targets set within the Action Programme, whenever they are called to decide over legislative proposals that result from the full or sectoral measurement that will be undertaken. • To explicitly share the EU SCM methodology, and commit to identify the IOs/DRs resulting from legislative initiatives and/or major amendments within the ex ante Impact Assessment procedure. • Not to approve any new proposal that is expected to increase administrative burdens. A good timing for this agreement would be March 2007, when the Council will issue its conclusions on the proposed Action Programme.140 4.3 Operational suggestions Apart from addressing methodological and organisational issues, this Pilot Project also dealt with operational issues, such as priority areas that could be selected for sectoral measurement and evidence of target-setting at national level. In this section we offer suggestions on how to use the results illustrated in Section 3 in the upcoming Action Programme. 4.3.1 Priority areas An important result of this Pilot Project is that policy areas identified by the four member states analysed coincide to a large extent. The areas that account for the lion’s share of administrative burdens in the surveyed member states are the following: • Annual accounts and company law; • Health Protection (including animal health and zootechnics) 140 It has been highlighted that the need to reach a wholly new Inter-Institutional Agreement would prove too ‘burdensome’ and lengthy. A possible alternative would imply that the Commission proposes an overall cost reduction target to be endorsed by the European Council in 2007. This, however, would not ensure the involvement of the European Parliament in pursuing the achievement of the announced reduction target. – 138 – • Working Environment and employment relations; • Fiscal Law and VAT; • Statistics; • Agriculture and agricultural subsidies; • Food labelling; and • Transport. The circumstance that national priority areas point at the same direction is important, but should be validated by the upcoming list of low-hanging fruits. As a matter of fact, gathering information about policy areas that account for the largest portion of administrative burdens is not sufficient to conclude that these areas are the ones that exhibit the largest reduction potential. Administrative burdens are, to be sure, not always superfluous, and even most burdensome areas might not lead to identifying suitable alternatives that would allow for abating administrative costs without compromising the effectiveness and efficiency of regulatory measures. However, if member states indicate low hanging fruits in the same areas, this would imply that these areas can be considered as candidates for successful sectoral measurements and cost reduction strategies. In this respect, we suggest that the Commission initially selects a sustainable number of policy areas (e.g. 8-10) that can be considered as particularly burdensome and characterised by a substantial reduction potential, and subjects them to sectoral measurement. The selected areas should be subjected to consultation in the preparatory phase of the Action Programme (October 2006-February 2007), to elicit proposals for simplification by member states and other stakeholders. As confirmed by available evidence in the four member states analysed, focusing on the ten most burdensome areas would mean, for the Commission, targeting at least 70% of the overall administrative burdens generated by EU and national legislation on businesses. Areas such as “Annual accounts and company law”, “Health Protection”, “Working Environment and employment relations” and “Fiscal Law and VAT” were found to create administrative burdens for businesses in Categories A, B and C. An in-depth sectoral measurement in these areas would likely allow the Commission to achieve results in a relatively short timeframe, provided that the measurement is organised effectively. Furthermore, for reasons already stated in the previous sections, we suggest that administrative costs imposed on businesses in remaining policy areas and administrative burdens for citizens and public administrations are addressed separately, possibly at a later stage. This suggestion is supported also by evidence reported in Table 4 above, showing that some of the member states that have not started a full-scale measurement have targeted some or all of these priority areas as particularly burdensome. Focusing on priority areas would, thus, allow the Commission to rely on a larger number of countries that are collecting data on administrative burdens in those areas. – 139 – 4.3.2 Target-setting Three of the four countries analysed have opted for setting an overall, ‘political’ target before undertaking the baseline measurement. The advantages of setting an overall target before the measurement are related to the important role that a strong political input plays in all better regulation initiatives. Not surprisingly, strong political input is included in the OECD principles of regulatory reform, as it creates a virtuous effect on accountability and ensures that monitoring and achievement of results takes place at all levels of government. At EU level, strong political commitment towards better regulation exists within the European Commission since many years. The extent to which an overall target can be set depends also on whether other EU institutions can commit to achieving a cost reduction target within a given timeframe. If inter-institutional agreement is reached on this issue, then we strongly recommend that the Commission announces a target in April 2007. As regards the magnitude of the target to be set, countries that have set an overall target have chosen a target of 25% (the Netherlands and Denmark) or a more conservative target of 20% (the Czech Republic). We suggest that the Commission decides on the magnitude of the target to be set after collecting simplification proposals during the consultation phase (October 2006-February 2007) and announces targets in April 2007. If the Commission receives a sufficient number of simplification proposals in the priority areas, then a 20% target could be attainable within a 5-year timeframe. An important issue is differentiating between overall targets and sectoral targets. If the Commission opts for a full-scale measurement in the Action Programme, then overall targets should be coupled by sectoral targets after the measurement has been carried out. These targets could be averaged at, e.g., 20%, but differentiated between involved DGs according to the estimated cost reduction potential (expressed in percentage). This is the approach that was adopted in the Netherlands and – although with an expected lower margin of differentiation – in the UK. However, we have suggested that the Commission award priorities to a selected number of policy areas when carrying out the measurement. Setting only sectoral targets ex ante is not inconsistent with the setting of an overall political target. As evidence gathered from the four member states analysed shows that the top 10 areas account for at least 70% of administrative burdens for businesses, setting a 25% reduction target in those areas would automatically imply that at least a 17.5% overall cost reduction target has been set. We also assume that the Action Programme will imply the setting of targets for member states. If this is the case, the magnitude of the target to be set is a delicate issue. It is important to note, in this respect, that the four countries analysed are not expected to belong to the group of EU countries with the highest percentage of administrative burdens on GDP. As confirmed by available evidence, such as the study carried out by Kox (2005) on 19 member states, the UK, Denmark, the Czech Republic and the Netherlands are all located below the average percentage found by the study. Table 36 below reports the findings of the study, which used extrapolation criteria – 140 – based on the results of a study by Djankov et al. (2002) on procedures of setting up a new firm in a selected sample of states. Table 36 – Estimated total administrative burden for EU countries, % of GDP, 2003 Source: Kox (2005) Based on these results, it would appear possible to set a 25% target for all member states. But further issues would need to be taken into account: • Some countries – e.g. the Netherlands and Denmark – have already undertaken efforts to reduce administrative burdens at national level. Imposing the same target for the years to come might thus prove discriminatory.141 • Some countries might not be able to count on sufficient budget resources to undertake a thorough measurement in the selected sectors. This also means that the 141One possibility would be that the Commission sets sectoral targets for member states, which can be reduced only if member states prove that they already achieved results in a specific area. This ‘reversal of the burden of proof’ would require member states to submit evidence and information, which would also lead to collecting best practices. – 141 – potential for implementing cost reduction strategies in these countries might be limited. Accordingly, we suggest that the Commission should set ex ante its target for reducing the administrative cost of EU legislation in the selected priority areas, and sets a more conservative target for member states within the given timeframe (e.g. 5 years). Alternatively, it was suggested that the Commission choose a target only for Categories A and B, then leaving it to Member States to decide whether to announce a target for national and regional legislation. However, we consider this solution to exhibit at least three problems: a) Need for commitment also at member state level: the Commission would still have to rely on actions taken by Member States, as Category B contains IOs/DRs falling under the sphere of influence of EU, international and national legislators: accordingly, announcing a target on Category B would still imply a similar commitment by member states. b) Weak political commitment: the overall target announced would not cover all administrative burdens faced by businesses in their everyday life, and would thus provide a much weaker signal in terms of commitment towards better regulation and competitiveness at pan-European level. Available evidence suggests that announcing a 25% reduction target on (all IOs/DRs included in) Categories A and B would imply committing to lift the burdens on businesses only by 10-14%; c) Different taxonomies: in light of the taxonomy adopted by the EU SCM tool, the Commission would rather have to announce a reduction target for Category ‘EU’. This, in turn, would imply that only the minimum implementation of EU pieces of legislation is included in the administrative burdens to be measured and subject to reduction proposals. In other words, the expected reduction target would probably represent less than 10% of total administrative burdens faced by EU businesses. 4.3.3 Periodic reviews v. ‘live’ baseline and the role of Impact Assessment Of the four member states analysed, the Netherlands and the Czech Republic chose to rely on periodic reviews of the baseline – i.e., a second round of measurement is foreseen. On the other hand, Denmark opted for yearly updates of the baseline. Finally, the UK Manual explains why the UK did not opt for a second measurement142: As the sample size is very small it is unlikely that any two samples drawn at different points in time would actually give comparable results. The point of a second measurement would be to compare results with the initial measurement. Therefore this makes conducting such an exercise of questionable value, especially given the high financial and resource costs of doing a measurement. Therefore the current plan for the UK is to carefully monitor the quality of the 142 See UK Manual, at p. 30. – 142 – flow data and use the database system to continually calculate the updated baseline to allow progress to be judged as described in the previous section. In addition there may be some scope for research to quality check the data. Another reason why the UK government opted for a ‘live’ baseline is that the UK – more than the other three countries analysed – has a consolidated RIA system since several years. With an established an pervasive RIA system, constantly updating the baseline simply requires that the SCM tool is fully integrated with the RIA. The European Commission has adopted a similar approach by integrating its operational manual in the IA guidelines and making the EU SCM operational at first in the ex ante impact assessment procedure. As a result, we consider that keeping a ‘live’ baseline would prove a suitable option at EU level, especially if the Commission decides to award priority to selected policy areas. The Commission is succeeding in integrating ex ante impact assessment in its Strategic Planning and Programming Cycle, requiring DGs to assess the economic, social and environmental impact of major new legislative proposals. If application of the EU SCM becomes pervasive in the Commission’s IA system, then keeping a ‘live’ baseline is a feasible option. If compared with the current model, the upcoming Action Programme would also require that Commission DGs identify the IOs/DRs introduced and/or repealed by new legislative proposals while performing the ex ante Impact Assessment. From this standpoint, the Secretariat General (or DG ENTR) should be awarded the power to require an identification of the IOs/DRs and an assessment of the corresponding cost: otherwise, the baseline would soon become obsolete. Importantly, if the Commission decides to maintain its current classification of origin, the identification of IOs/DRs in impact assessment should also include an assessment of the ‘minimum level of implementation’, both for proposed directives and – to a lesser extent – proposed regulations. 4.4 Summary of main findings In this section, we have identified some suggestions for the upcoming Commission Action Programme, with specific respect to the practical implementation of the EU SCM in combination with existing and future national SCM variants. In particular, we reached the following findings. • Based on national experiences and the similarities between the EU SCM and national SCM variants, launching a pan-European Action Programme is feasible, although challenging. • The issue of database interoperability and institutional communication between the EU and member states would be more easily solved if the Commission identifies all the IOs/DRs generated by the acquis, plus those that represent the ‘minimum efficient implementation’ of pieces of EU legislation. This activity – which leads to defining the ‘EU’ category (or category ’zero’) should be carried out by consultants in the first months of the Action Programme. Once the Commission has published a list of IOs/DRs, member states that already have built a database should update – 143 – it accordingly. Those that do not have a database should design it in order to allow for interoperability – i.e. by ensuring that all the IOs defined by the Commission are flagged and measured. • With this arrangement, member states will be called to validate Category ‘EU’ and then define their category B in a residual way, thus also identifying instances of gold-plating. • Member states are more suited to measure IOs/DRs provided by the Commission as ‘EU’. As most member states will have to carry out their own measurement anyway, such an arrangement increases economies of scope and avoids redundancies such as, e.g., double interviews or duplication of focus groups, etc. Depending on the number of countries that participate to the measurement, the Commission will then have to use a country distribution to extrapolate the results to the EU level. Such exercise has already been undertaken in recent Commission RIAs that integrate the SCM tool. • If the Commission defines category ‘EU’, there would be no need to invest efforts and resources to build a ‘table of concordance’ between EU and national legislation during the preparatory phase of the measurement. At the end of the measurement, if member states have flagged the origin of – and the national references for – all IOs defined by the Commission as ‘EU’, the table of concordance would be automatically generated. • Assumptions made during the Action Programme and in extrapolating data to the EU level can then be standardised for use in ex ante impact assessment, though the SCM methodology for ex ante impact assessment (like occurs in the UK) will inevitably differ from that applied in the ex post measurement exercise. In ex ante measurement, as a matter of fact, one-off costs should always be taken into account, and ‘business as usual’ costs should never be considered. • As regards accuracy and the level of detail, we suggest – based on national experiences analysed – that the Commission identifies ‘IOs’ as the basic unit of its measurement, though subdividing IOs into DRs only in exceptional circumstance. Further subdivisions would add little accuracy, but a lot of complexity. Furthermore, the typical ‘Pareto distribution’ of administrative burdens suggests that in most cases, for IOs that are not particularly costly direct assessment of administrative cost is more efficient and effective than use of more costly and time- consuming data collection tools. • In operational terms, we suggest that the Commission awards priority to the 10 most burdensome areas in Categories A and B, submitting them to consultation in the October 2006-February 2007 timeframe. These areas represent at least 70% of burdens generated by EU legislation. The measurement of remaining areas and of burdens faced by citizens and public administrations should be undertaken at a later stage or, in any case, separately. – 144 – • As the EU impact assessment system is becoming more established and pervasive in the Commission Strategic Planning and Programming Cycle, we suggest that the Commission opts for keeping a ‘live’ baseline. This requires some refinement in the current IA procedure: in particular, when performing ex ante impact assessments, DGs, the Parliament and the Council should ensure that the IOs corresponding to new proposals are clearly identified, and that IOs that are repealed are also listed. • Targets can be set for priority areas. This does not preclude the potential for announcing an overall ‘political’ target, as was done in some of the analysed countries. After the measurement, targets can be adjusted for each DG. • When setting targets, it is essential that other EU institutions commit to cooperate with the Commission in achieving the targets set. Otherwise, simply setting targets for each DG would prove impossible. Such a coordination between the Commission, the Council and the Parliament should be sought through an Inter- Institutional Agreement. • Institutional communication between the Commission an member states can be achieved by replicating the fruitful collaboration with the International SCM Network that made this Final Report possible. As the participation of EU member states to the Network is increasing, this would automatically ensure that the Commission avails of a growing number of counterparts and data sources when undertaking the measurement. Moreover, the use of web-based tools such as SINAPSE would prove essential.